The Effects of Minimum Caseload Requirements on Management and Outcome in Abdominal Aortic Aneurysm Repair
A secondary analysis of German DRG statistics data
; ; ; ; ;
Background: The German quality assurance guideline on abdominal aortic aneurysm (AAA) was implemented by the Joint Federal Committee (Gemeinsamer Bundesausschuss, G-BA) in 2008. The aims of this study were to verify the association between hospital case volume and outcome and to assess the hypothetical effect of minimum caseload requirements.
Methods: The German diagnosis-related groups statistics for the years 2012 to 2016 were scrutinized for AAA (ICD-10 GM I71.3/4) with procedure codes for endovascular or open surgical treatment. The primary endpoint was in-hospital mortality. Logistic regression models were used for risk adjustment, and odds ratios (OR) were calculated as a function of the annual hospital-level case volume of AAA. In a hypothetical approach, the linear distances for various minimum caseloads (MC) were evaluated to assess accessibility.
Results: The mortality of intact AAA (iAAA) was 2.7% (men [M] 2.4%, women [W] 4.2%); ruptured AAA (rAAA), 36.9% (M 36.9%, F 37.5%). An inverse relationship between annual hospital case volume of AAA and mortality was confirmed (iAAA/rAAA: from 3.9%/51% [<10 cases/year] through 3.3%/37% [30–39 cases/year] to 1.9%/28% [≥ 75 cases/year]). For a reference category of 30 AAA procedures/year, the following significant OR were found: 10 AAA cases/year, OR 1.21 (95% confidence interval [1.20; 1.21]); 20 cases, OR 1.09 [1.09; 1.09]; 50 cases, OR 0.89 [0.89; 0.89]; 75 cases, OR 0.82 [0.82; 0.82]. In a hypothetical centralization scenario with assumed MC of 30/year, 86% of the population would have to travel less than 100 km to the nearest hospital; with an MC of 40, this would apply to only 50% (without redistribution effects).
Conclusion: In the observed period, a significant correlation was confirmed between high annual case volume and low in-hospital mortality. A minimum caseload requirement of 30 AAA operations/year seems reasonable in view of the accessibility of hospitals.
Abdominal aortic aneurysms (AAA) are localized dilatations of the infrarenal aorta. The rupture rate of AAA with a maximal diameter (MD) <5.5 cm is less than 1.5% for male patients (females <6%), while the rupture rate for AAA with an MD of >7.0 cm exceeds 20% (females >40%) (1). Ruptured AAA—with or without surgical treatment—bears a high risk of mortality, generally >80% (2, 3). To prevent AAA rupture, prophylactic repair is recommended in the presence of certain clinical and morphological criteria, e.g., maximal aneurysm diameter >50–55 mm or rapid diameter progression (>10 mm/year) (4, 5, 6, 7). Depending on surgical risk, life expectancy, and comorbidities, treatment is typically endovascular (endovascular aortic repair, EVAR) or open surgery (open aortic repair, OAR). Current recommendations for diagnosis, treatment, and follow-up of AAA can be found in the national guideline published in July 2018 (5).
Diagnosis, indication to treat, and performance of these interventions require a high degree of expertise. To improve and ensure quality of AAA surgery, effectual directives were implemented by the Federal Joint Committee (Gemeinsamer Bundesausschuss, G-BA) in 2008. Hospitals treating AAA are, for example, obliged to offer a vascular on-call service 24/7, AAA operations must be performed or assisted by specialist vascular surgeons, and strict rules apply to qualification of intensive care unit staff (8). Despite a high level of evidence for an association between high annual hospital volume and low mortality, a minimal threshold of annual AAA caseload was not included in the above-mentioned AAA directive (9, 10, 11). As recent German studies have included years preceding implementation of the G-BA measures, the current validity of their results is to be questioned (12, 13).
Therefore, the aim of the present study was to reanalyze the volume–outcome effect in Germany on the basis of the most recent diagnosis-related groups (DRG) data available from the German Federal Statistical Office (GFSO; up to 2016) and to assess the effect of hypothetical minimum volumes on geographical accessibility of hospitals treating AAA.
Since all DRG data have to be submitted to the GFSO, this analysis can be considered a full survey of the German population. The study complies with the guidelines and recommendations for good practice of secondary data analysis (15), and the manuscript was written in accordance with STROSA2 (German Standardized Reporting Routine for Secondary Data Analysis) (16).
Data from the GFSO and Federal Institute for Research on Building, Urban Affairs, and Spatial Development (BBSR; INKAR database) were linked for calculation of linear distance between a patient’s residence and the geocoded hospital location.
The administrative codes used in this analysis are listed in eTable 1. Generally, all DRG cases with a diagnosis of AAA (ICD-10-GM: I71.4 = intact AAA [iAAA]; I71.3 = ruptured AAA[rAAA]) and German operation procedure codes (OPS) for OAR or EVAR were included.
Patient and hospital characteristics, outcomes
The most recent available patient and treatment characteristics as well as the outcomes for 2016 are reported.
A volume–outcome analysis was conducted for the most recent 5-year period (2012–2016). Hospitals were grouped according to their annual caseload of AAA treated by open and endovascular means. Patient characteristics, treatment modality, and outcomes (primary endpoint: in-hospital mortality) are described. The analysis was adjusted for patient risk. eFigure 1 shows a patient flow chart.
For hospital-level characterization of patients, the median as well as the quartiles Q1 and Q3 are reported to describe the variance within the respective hospital volume group. For volume–outcome analysis, odds ratios (OR) for in-hospital mortality after iAAA and rAAA repair were calculated as a function of annual volume (modeled as continuous variable).
In a hypothetical approach, the linear distance to the closest hospital fulfilling the arbitrarily set caseload requirement (minimal caseload, MC = 10/20/30/40/50 AAA cases/year) was calculated for the whole population of Germany.
Status quo: patient characteristics, treatment, and outcome in 2016
The characteristics, comorbidities, and details of the management, treatment, and outcomes of 11 931 patients receiving treatment for AAA (89.6% iAAA) in 475 hospitals are presented in Table 1 and eTable 2. The median age was 74 [Q1; Q3: 66; 79] years (iAAA 73 [66; 78]; rAAA 75 [67; 81] years), and 86.5% of patients were men (iAAA 86.8%; rAAA 85.2%).
EVAR was performed in 75.8% of male patients (iAAA 79.8%; rAAA 40.7%) and in 70.8% of females (iAAA 73.8%; rAAA 47.3%; Table 1).
The complication rates were generally higher in rAAA than iAAA. The length of stay after iAAA repair was 9 (7; 14) days, after rAAA repair 13 (6; 25) days. In-hospital mortality was 2.4% for men receiving treatment for iAAA and 4.2% for women. Treatment of rAAA showed mortality rates of 36.9% and 37.5%, respectively (Table 1).
The median linear distance from a patient’s residence to the hospital was 11.3 (5.3; 23.5) km. There was no relevant difference in distance between iAAA and rAAA (Table 1).
Regarding hospital volume, 4.4% (530 cases) were treated in hospitals that operated on <10 AAA in 2016; 65.1% (7766 cases) and 37.5% (4465 cases) were treated in hospitals operating on >30 and >50 cases, respectively (eTable 2).
A contingency table of the district type of patient’s residence and the location of the treating hospital is shown in eTable 3. Most patients (3193 cases) lived and were treated in administratively independent cities. eFigure 2 shows the share of patients treated per share of hospitals in 2016 (Lorenz curve).
Volume–outcome analysis 2012 to 2016
Hospital-level characteristics and outcomes according to hospital volume groups for the accumulated period from 2012 to 2016 are listed in Table 2. The overall number of hospitals treating AAA (2012: 507; 2016: 475) and the distribution into volume groups for each year separately are shown in eFigure 3.
Patient-level outcomes in relation to annual hospital volume are given in Table 3. In-hospital mortality of patients treated in hospitals with <10 AAA operations/year was highest (iAAA 3.9%; rAAA 51.3%), while it was lowest in hospitals treating ≥75 cases/year (iAAA 1.9%; rAAA 27.8%) (Figure 1, eFigure 4).
The volume–outcome relationship is depicted in Figure 2. Volume was modeled as a continuous variable using multilevel multivariable regression analysis (Figure 2a). For Figure 2b, a multivariable multilevel regression model was used to calculate respective mortality odds, with all volumes between 2 and 100 modeled as a dichotomous variable. For every assumed annual caseload within that range, the risk of mortality is higher in hospitals with a caseload below any given number, compared to hospitals with a caseload equal to or higher than that number. This is evident as the odds ratios and confidence intervals are all larger than 1.
Hypothetical effects of minimal caseload requirements
Figure 3 shows a hypothetical model (“what-if” scenario) of geographical hospital accessibility for the whole German population with arbitrarily set minimal caseload requirements. This model shows how far patients would have to travel to reach the next hospital with an annual AAA caseload of at least X. With an assumed MC of 30 AAA procedures/year, 86% of the population would have to travel <100 km to the nearest hospital (98% <150 km). For an MC = 40, this would apply to only 50% (69% <150 km). Redistribution effects could not be considered or modeled due to data protection restrictions.
Following the G-BA’s implementation of quality assurance measures in 2008, AAA are still treated in 475 hospitals in Germany. The inverse association of AAA annual operative caseload and in-hospital mortality persisted during the observation period (2012–2016). If a minimal case threshold were to be implemented, the population’s access to hospitals would best be preserved by choosing a threshold of 30 AAA procedures/year.
The overall mortality of iAAA in 2016 was 2.4% for men and 4.2% for women; for rAAA 36.9% and 37.5% respectively. Comparing these rates to an earlier analysis up to 2014 (3), mortality seems to be relatively stable on this level; it decreased, however, in the years before the present analysis.
Volume–outcome effect and centralization of AAA treatments
In 2016, 475 hospitals treated AAA and 63% of all AAA were treated in 416 hospitals with an annual caseload <50 (iAAA 62%; rAAA 65%). The 2011 European AAA treatment guidelines recommended repair only in hospitals with a caseload of >50 elective AAA procedures/year (18). These guidelines were updated in 2019, and AAA repair is now recommended only in hospitals with >30 annual cases (7). This guideline also advises against repair in hospitals with <20 AAA operations/year. In 2016 a caseload of <30 was evident in 327 hospitals in Germany, in which 34.9% of all procedures were carried out (iAAA 35%; rAAA 34.8%). A further 17.3% of cases received treatment in 241 hospitals with <20 annual operations (iAAA 17.1%, rAAA 19%). The volume–outcome effect for AAA repair has been shown for many countries and healthcare systems (13, 19, 20, 21, 22, 23) and was confirmed in the present study.
This is the first study on volume–outcome effects in AAA surgery focusing solely on years after the implementation by the G-BA of quality assurance directives for AAA surgery in Germany in 2008. The G-BA recommendations comprised several measures related to a surgeon’s vascular surgery qualification, staffing in intensive care units, etc. (8). Despite these structural quality measures, the effect of annual hospital volume on in-hospital mortality in AAA repair remained evident. In-hospital mortality was highest in low-volume hospitals and lowest in hospitals treating >75 AAA/year. The patient-level raw data (Table 3, Figure 1) show this inverse relationship between annual caseload and mortality. In contrast, the raw data presented in Table 2 and eFigure 4 refer to hospital level. Mortality as well as other characteristics at “hospital level” are expressed as median with first and third quartiles. This permits assessment of the variance of patient cohorts between the hospital groups. As the distribution of hospital- specific mortality is right-skewed especially in the low-volume groups, the median is considered more valid than the arithmetic mean.
In the multivariable model, in which volume was modeled as a continuous variable, a possible risk reduction seems to persist over the entire scale (up to 100 cases per year). However, the aim of this analysis was not the biostatistical calculation of “the one and only optimal” threshold, but a hypothetical assessment: what if the minimum caseload requirement were set to a specific value. As shown by the steadily descending curve in Figure 2a, every threshold would result in significant higher mortality in low-volume versus high-volume centers. This is underlined by the data presented in Figure 2b. Here, the lower limits of confidence intervals are always greater than 1.
Looking at the travel distances for patients, selecting 75 cases/year as the minimum caseload seems unrealistic. With regard to the distance from hospital alone, there is a large gap between MC30 and MC40. It thus seems that a minimum caseload 30 cases might be more reasonable. In this scenario, 86% of patients would have to travel less than 100 km to the nearest hospital and 98% would have a hospital within 150 km. It is also assumed that the effects of redistributing patients away from smaller hospitals, together with structural hospital planning, would lead to even smaller travel distances. Due to data protection issues, we could not model these redistribution effects. However, the increase in travel burden with the hypothetical minimum caseloads seen in our analysis is significant. This is in line with a previous report in 2015 (24). Moreover, a threshold of 30 AAA cases/year would be congruent with the above-mentioned European guideline (7). Finally, a recent paper on volume–outcome relationships in Germany showed lower in-hospital mortality for the fourth quartile (>30 cases/year) than for all other quartiles (13).
It is often discussed whether treatment for rAAA—an emergency that requires immediate care—can be assured in a scenario of centralization. A literature review published together with an analysis of rAAA travel distances that took deaths on the way to the hospital and in the community into account, did not show an association between distance traveled (as a stand-alone variable) and outcome (25).
Future training of vascular surgeons might also benefit from centralization provided the latter is organized centrally (as for example in the UK). Currently, open aortic operations are part of the vascular surgeon’s training curriculum. However, adequate training in these complex operations can only be conducted in hospitals with a high caseload (26). With almost 80% of cases managed endovascularly, a general consensus on future vascular training may well be indispensable.
As in earlier studies, only AAA cases admitted to a hospital and treated with EVAR or OAR were available for the present analysis. The number of undetected/untreated AAA remains unclear. Especially patients with rAAA who were found dead at home, died before arrival of an ambulance or being admitted to a hospital, or died before being correctly diagnosed as rAAA (e.g., by ultrasound or computed tomography) are missing for analysis. These data are not collected in Germany; cause of death statistics refer only to the underlying disease. As the autopsy rate is low in Germany, we suspect that many deaths due to rAAA could be coded otherwise. Nevertheless, the effect of the time that elapses between AAA rupture and hospital admission on mortality or secondary outcomes is of great interest for further discussions on centralization of vascular surgery services.
Some AAA may have been coded as AA of unspecific origin (I71.8/9). As it is not possible to determine the location by DRG coding for these codes, they were excluded from analysis.
As DRG data were used, in which only hospital mortality is recorded, no analysis of of the course post discharge was possible. Another source of bias could be incorrect coding, although due to the large number of cases this should have little effect on the overall survey. Earlier studies showed an association of surgeon’s annual case volume with postoperative outcome (29, 30); however, DRG data include no surgeon-level information, so this could not be analyzed. Owing to data protection laws, only the geographical centroid of the municipality where each patient resided could be used and the effects of secondary distribution could not be analyzed.
We thank all employees of the GFSO who supported our research, specifically Melanie Scheller, Jutta Spindler, and Sabine Nemitz.
Conflict of interest statement
Dr. Trenner has received reimbursement of travel costs and congress attendance fees from Medtronic, Gore, Cook Medical, Bolton Medical und Terumo Aortic; payments for lectures from Terumo Aortic; and study support (third-party funding) from Medtronic.
PD Dr. Busch acts as clinical consultant for Brainlab AG. He has received a lecture fee from Terumo Aortic and study support (third-party funding) from the German Heart Foundation (Deutsche Herzstiftung).
Prof. Eckstein has received study support (third-party funding) from Medtronic.
The remaining authors declare that no conflict of interest exists.
Manuscript received on 3 April 2020, revised version accepted on 3 September 2020
Prof. Dr. med. Andreas Kühnl, MPH, MBA
Klinik und Poliklinik für Vaskuläre und Endovaskuläre Chirurgie
Klinikum rechts der Isar der Technischen Universität München
Ismaninger Str. 22, 81675 München, Germany
Cite this as:
Trenner M, Salvermoser M, Busch A, Schmid V, Eckstein HH, Kühnl A:
The effects of minimum caseload requirements on management and outcome in
abdominal aortic aneurysm repair—a secondary analysis of German DRG statistics
data. Dtsch Arztebl Int 2020; 117: 820–7. DOI: 10.3238/arztebl.2020.0820
►Supplementary material to:
eMethods, eTables, eFigures:
Department of Statistics, Ludwig Maximilians University Munich: Prof. Dr. Volker Schmid
|1.||Grant SW, Sperrin M, Carlson E, et al.: Calculating when elective abdominal aortic aneurysm repair improves survival for individual patients: development of the Aneurysm Repair Decision Aid and economic evaluation. Health Technol Assess 2015; 19: 1–154 CrossRef MEDLINE PubMed Central|
|2.||Dueck AD, Kucey DS, Johnston KW, Alter D, Laupacis A: Survival after ruptured abdominal aortic aneurysm: effect of patient, surgeon, and hospital factors. J Vasc Surg 2004; 39: 1253–60 CrossRef MEDLINE|
|3.||Kühnl A, Erk A, Trenner M, Salvermoser M, Schmid V, Eckstein HH: Incidence, treatment and mortality in patients with abdominal aortic aneurysms—an analysis of hospital discharge data from 2005–2014. Dtsch Arztebl Int 2017; 114: 391–8 VOLLTEXT|
|4.||Chaikof EL, Dalman RL, Eskandari MK, et al.: The Society for Vascular Surgery practice guidelines on the care of patients with an abdominal aortic aneurysm. J Vasc Surg 2018; 67: 2–77 CrossRef MEDLINE|
|5.||Debus ES, Heidemann F, Gross-Fengels W, et al.: [Short version of the S3 guideline on screening, diagnosis, therapy and follow-up of abdominal aortic aneurysms]. Gefasschirurgie 2018; 23: 432–51 CrossRef|
|6.||Erbel R, Aboyans V, Boileau C, et al.: 2014 ESC guidelines on the diagnosis and treatment of aortic diseases: document covering acute and chronic aortic diseases of the thoracic and abdominal aorta of the adult. The Task Force for the Diagnosis and Treatment of Aortic Diseases of the European Society of Cardiology (ESC). Eur Heart J 2014; 35: 2873–926 CrossRef MEDLINE|
|7.||Wanhainen A, Verzini F, Van Herzeele I, et al.: Editor‘s Choice – European Society for Vascular Surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal Aorto-iliac Artery Aneurysms. Eur J Vasc Endovasc Surg 2019; 57: 8–93 CrossRef MEDLINE|
|8.||Gemeinsamer Bundesausschuss: Richtlinie über Maßnahmen zur Qualitätssicherung für die stationäre Versorgung bei der Indikation Bauchaortenaneurysma – QBAA-RL, Berlin 2020. www.g-ba.de/richtlinien/65/ (last accessed on 21 October 2020).|
|9.||Holt PJ, Poloniecki JD, Gerrard D, Loftus IM, Thompson MM: Meta-analysis and systematic review of the relationship between volume and outcome in abdominal aortic aneurysm surgery. Br J Surg 2007; 94: 395–403 CrossRef MEDLINE|
|10.||Karthikesalingam A, Hinchliffe RJ, Loftus IM, Thompson MM, Holt PJ: Volume-outcome relationships in vascular surgery: the current status. J Endovasc Ther 2010; 17: 356–65 CrossRef MEDLINE|
|11.||Phillips P, Poku E, Essat M, et al.: Procedure volume and the association with short-term mortality following abdominal aortic aneurysm repair in European populations: a systematic review. Eur J Vasc Endovasc Surg 2017; 53: 77–88 CrossRef MEDLINE|
|12.||Trenner M, Haller B, Söllner H, et al.: Twelve years of the quality assurance registry on ruptured and non-ruptured abdominal aortic aneurysms of the German Vascular Society (DGG)—part 3: predictors of perioperative outcome with a focus on annual caseload. Gefasschirurgie 2015; 20 (S1): 32–44 CrossRef|
|13.||Trenner M, Kuehnl A, Salvermoser M, et al.: Editor‘s Choice: High annual hospital volume is associated with decreased in hospital mortality and complication rates following treatment of abdominal aortic aneurysms: secondary data analysis of the nationwide German DRG Statistics from 2005 to 2013. Eur J Vasc Endovasc Surg 2018; 55: 185–94 CrossRef MEDLINE|
|14.||FDZ der Statistischen Ämter des Bundes und der Länder: Fallpauschalenbezogene Krankenhausstatistik (Research data centers of the federal statistical office: DRG statistics) eigene Berechnungen 2005–2016.|
|15.||Swart E, Gothe H, Geyer S, et al.: Good Practice of Secondary Data Analysis (GPS): guidelines and recommendations. Gesundheitswesen 2015; 77: 120–6 CrossRef MEDLINE|
|16.||Swart E, Bitzer EM, Gothe H, et al.: STandardisierte BerichtsROutine für Sekundärdaten Analysen (STROSA) – ein konsentierter Berichtsstandard für Deutschland, Version 2. Gesundheitswesen 2016; 78 (S 01): e145–e160; 161 CrossRef MEDLINE|
|17.||Wilson EB: Probable inference, the law of succession, and statistical inference. J Am Stat Assoc 1927; 22: 209–12 CrossRef|
|18.||Moll FL, Powell JT, Fraedrich G, et al.: Management of abdominal aortic aneurysms clinical practice guidelines of the European Society for Vascular Surgery. Eur J Vasc Endovasc Surg 2011; 41 (S 1): S1–58 CrossRef MEDLINE|
|19.||Holt PJ, Karthikesalingam A, Hofman D, et al.: Provider volume and long-term outcome after elective abdominal aortic aneurysm repair. Brit J Surg 2012; 99: 666–72 CrossRef MEDLINE|
|20.||Landon BE, O‘Malley AJ, Giles K, Cotterill P, Schermerhorn ML: Volume-outcome relationships and abdominal aortic aneurysm repair. Circulation 2010; 122: 1290–7 CrossRef MEDLINE|
|21.||Scali ST, Beck AW, Sedrakyan A, et al.: Hospital volume association with abdominal aortic aneurysm repair mortality: analysis of the International Consortium of Vascular Registries. Circulation 2019; 140: 1285–7 CrossRef MEDLINE|
|22.||Trenner M, Salvermoser M, Busch A, Reutersberg B, Eckstein HH, Kuehnl A: Effect modification of sex and age for the hospital volume-outcome relationship in abdominal aortic aneurysm treatment: secondary data analysis of the nationwide German diagnosis related groups statistics from 2005 to 2014. J Am Heart Assoc 2020; 9: e014534 CrossRef MEDLINE PubMed Central|
|23.||Zettervall SL, Schermerhorn ML, Soden PA, et al.: The effect of surgeon and hospital volume on mortality after open and endovascular repair of abdominal aortic aneurysms. J Vasc Surg 2017; 65: 626–34 CrossRef CrossRef PubMed Central|
|24.||Hentschker C, Mennicken R: The volume-outcome relationship and minimum volume standards–empirical evidence for Germany. Health Econ 2015; 24: 644–58 CrossRef MEDLINE|
|25.||Souza VC, Strachan DP: Relationship between travel time to the nearest hospital and survival from ruptured abdominal aortic aneurysms: record linkage study. J Public Health (Oxf) 2005; 27: 165–70 CrossRef MEDLINE|
|26.||Smith ME, Andraska EA, Sutzko DC, Boniakowski AM, Coleman DM, Osborne NH: The decline of open abdominal aortic aneurysm surgery among individual training programs and vascular surgery trainees. J Vasc Surg 2020; 71: 1371–7 CrossRef MEDLINE|
|27.||Trenner M, Kuehnl A, Reutersberg B, Salvermoser M, Eckstein HH: Nationwide analysis of risk factors for in-hospital mortality in patients undergoing abdominal aortic aneurysm repair. Brit J Surg 2018; 105: 379–87 CrossRef MEDLINE|
|28.||Trenner M, Eckstein HH, Kallmayer MA, Reutersberg B, Kühnl A: [Secondary analysis of statutorily collected routine data]. Gefasschirurgie 2019; 24: 220–7 CrossRef|
|29.||Marlow NE, Barraclough B, Collier NA, et al.: Effect of hospital and surgeon volume on patient outcomes following treatment of abdominal aortic aneurysms: a systematic review. Eur J Vasc Endovasc Surg 2010; 40: 572–9 CrossRef MEDLINE|
|30.||McPhee JT, Robinson WP, 3rd, Eslami MH, Arous EJ, Messina LM, Schanzer A: Surgeon case volume, not institution case volume, is the primary determinant of in-hospital mortality after elective open abdominal aortic aneurysm repair. J Vasc Surg 2010; 53: 591–9.e2 CrossRef MEDLINE|
|31.||Benchimol EI, Smeeth L, Guttmann A, et al.: The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Med 2015; 12: e1001885 CrossRef MEDLINE PubMed Central|
|32.||Krautz C, Nimptsch U, Weber GF, Mansky T, Grützmann R: Effect of hospital volume on in-hospital morbidity and mortality following pancreatic surgery in Germany. Ann Surg 2018; 267: 411–7 CrossRef MEDLINE|
|33.||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|
|34.||Nimptsch U, Krautz C, Weber GF, Mansky T, Grützmann R: Nationwide in-hospital mortality following pancreatic surgery in Germany is higher than anticipated. Ann Surg 2016; 264: 1082–90 CrossRef MEDLINE|
|35.||Nimptsch U, Mansky T: Trends in acute inpatient stroke care in Germany—an observational study using administrative hospital data from 2005–2010. Dtsch Arztebl Int 2012; 109: 885−92 CrossRef MEDLINE PubMed Central|
|36.||Quan H, Sundararajan V, Halfon P, et al.: Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005; 43: 1130–9 CrossRef MEDLINE|
|37.||van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ: A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care 2009; 47: 626–33 CrossRef MEDLINE|
|38.||Vincenty T: Direct and inverse solutions of geodesics on the ellipsoid with application of nested equations. Surv Rev 1975; 23: 88–93 CrossRef|