Hip and Knee Replacement in Germany and the USA
Analysis of Individual Inpatient Data From German and US Hospitals for the Years 2005 to 2011
Background: The number of hip and knee replacement operations is rising in many industrialized countries. To evaluate the current situation in Germany, we analyzed the frequency of procedures in Germany compared to the USA, with the aid of similar case definitions and taking demographic differences into account.
Methods: We used individual inpatient data from Germany (DRG statistics) and the USA (Nationwide Inpatient Sample) to study differences in the age- and sex-adjusted rates of hip and knee replacement surgery and the determinants of trends in case numbers over the years 2005 to 2011.
Results: In 2011, hip replacement surgery was performed 1.4 times as frequently in Germany as in the USA (284 vs. 204 cases per 100 000 population per year; the American figures have been adjusted to the age and sex structure of the German population). On the other hand, knee replacement surgery was performed 1.5 times as frequently in the USA as in Germany (304 [standardized] vs. 206 cases per 100 000 population per year). Over the period of observation, the rates of both procedures increased in both countries. The number of elective primary hip replacement operations in Germany grew by 11%, from 140 000 to 155 300 (from 170 to 190 per 100 000 persons); after correction for demographic changes, a 3% increase remained. At the same time, the rate of elective primary hip replacement surgery in the USA rose by 28%, from 79 to 96 per 100 000 population, with a 13% increase remaining after correction for demographic changes.
Conclusion: There are major differences between Germany and the USA in the frequency of these operations. The observed upward trend in elective primary hip replacement operations was mostly due to demographic changes in Germany; non-demographic factors exerted a stronger influence in the USA than in Germany. With respect to primary knee replacement surgery, non-demographic factors exerted a comparably strong influence in both countries.
The increase in numbers of hip and knee replacements and the frequency of such surgery in comparison to other countries is the subject of critical discussion in Germany. The overriding impression is that such surgery is performed relatively frequently in Germany when compared to other countries (1–3). Most of the published data on this subject is based on the same primary source, namely OECD indicators (4).
According to OECD reports, 286 hip replacement operations were performed per 100 000 population in Germany in 2011. This places Germany in second place among OECD countries for hip replacement frequency, behind Switzerland. Germany lies in third place for frequency of knee replacements, with 207 operations per 100 000 population (according to OECD figures), behind the USA and Austria (5). Previous years’ OECD reports also state high surgery frequencies for Germany (4, 6, 7). However, the extent to which such comparisons can be interpreted is limited, partly because they do not take into account differing demographics and partly because numbers of surgeries are calculated in different ways in different countries.
Unlike OECD reports, this article is not based on statistics prepared by others. Instead, it uses individual inpatient data to compare hip and knee replacements in Germany with those in the USA. Formally and qualitatively comparable patient data from these two countries is available and accessible, so that clinical entities can be precisely defined and evaluated on the basis of individual cases. This article analyzes demographically adjusted differences in frequency and determining factors behind changes in case numbers during the observation period.
For Germany, nationwide inpatient statistics (DRG statistics), including treatment data on all inpatient cases processed according to the DRG system, were evaluated (8). In 2011, DRG statistics covered approximately 17.7 million patients in 1600 hospitals.
For the USA, the Nationwide Inpatient Sample (NIS) was used. This contains the data on all inpatients in a representative sample of 20% of US hospitals (9), providing information on approximately 8 million inpatients for each year of the observation period in 1000 US hospitals.
The unit of analysis is an inpatient who underwent hip or knee replacement surgery.
In Germany, surgical procedures are coded according to the Surgery and Procedure Coding System (OPS, Operationen- und Prozedurenschlüssel), and diagnoses are coded according to ICD-10-GM (International Classification of Diseases, Tenth Revision, German Modification). In the USA, ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification) is used; this includes both diagnoses and procedures.
Case definitions are based on the inclusion criteria for documenting hip and knee replacements according to statutory quality assurance (10) but have been modified, as different classification systems include different levels of detail (eTable 1). These definitions, which provide comparable information on joint replacement surgery in Germany and the USA, were used in the same way for each year of the observation period; this means that longitudinal comparisons are also possible. The definition of hip replacement includes both total and partial joint replacement and is divided into elective primary replacement, primary replacement for fracture, revision replacement, and replacement for other indications. The definition of knee replacement includes both total and partial joint replacement (excluding isolated patella replacement) and is divided into primary replacement and revision replacement. All included patients were aged 20 years and older.
Surgery frequencies were reported on an annual basis for both countries. Because the US data was obtained from a sample, national frequencies were estimated on the basis of the stratified weighting factors stated in the NIS dataset (9). Crude rates per 100 000 population are given in addition to absolute frequencies. Crude rates were calculated by dividing case numbers by the total population for the year (11–13). To enable comparisons to be made between the two countries, annual surgery rates for the USA, standardized for sex and age to match German demographics, were calculated (direct standardization by sex and five-year age groups for each year of the observation period).
Changes over time were analyzed using multiplicative decomposition of the Laspeyres index (eBox 1). This includes aspects of changes in case numbers between 2005 and 2011 that were determined by demographics, as well as those that were independent of demographics (14, 15).
The demographics-related changes reported in this way can be ascribed to shifts in demographics such as population aging or growth. Changes that were independent of demographics (as shown in figures standardized for age and sex) are the result of other factors affecting the frequency of surgery.
Demographic parameters, crude surgery frequencies
The German and US populations changed in different ways during the observation period. While the US population grew by 5%, the German population fell by 1%. The proportion of those aged over 65 increased in both countries. In 2011 this figure was 21% for Germany and 13% for the USA.
Hip replacement frequency increased in both countries (Table 1). The crude rate per 100 000 population increased from 254 to 284 in Germany and from 129 to 149 in the USA. In terms of indication, in Germany approximately two-thirds of operations were elective primary replacements, 21% were for fracture, and 10% were revision replacements. Distribution in the USA was similar.
Knee replacements also increased during the observation period: from 164 to 206 operations per 100 000 population in Germany and from 181 to 225 per 100 000 population in the USA. In Germany the proportion of revision replacements increased from 7.2% in 2005 to 9.5% in 2011; in the USA it rose from 7.4% to 8.4%.
The proportion of patients aged over 65 was higher in Germany than in the USA for both hip and knee replacements (Table 1).
Comparison standardized for sex and age
Demographics are only part of the reason hip replacement rates are higher in Germany than in the USA. While the crude (i.e. not adjusted for demographics) rate in 2011 was 149 per 100 000 population in the USA, when standardized to match German demographics it was 204. However, with 284 surgeries per 100 000 population, hip replacements were approximately 1.4 times more frequent in Germany even after adjustment for demographic differences.
This difference, which can be seen in all years of the observation period, holds true for primary replacements for fracture and revision replacements as well as for elective primary replacements, although the latter is the most significant indication numerically (Figure 1a, eTable 2). Figure 2a shows age-specific rates for primary replacements. The difference was particularly marked in the 70 to 79 age groups: here the German rates were almost twice the US rates.
For knee replacements, the crude rates in the USA were higher than those in Germany. This difference increased further once the US figures were standardized for sex and age to match the German population: in 2011 the standardized rate for the USA was 304 operations per 100 000 population, versus 206 in Germany. Knee replacement was thus 1.5 times more frequent in the USA than in Germany after adjustment for demographic differences. This difference in frequency was present in all years of the observation period (Figure 1b, eTable 3). Figure 2b shows that the frequency of primary replacements was higher in the USA than in Germany in almost every age group.
Changes in case numbers
The number of hip replacements increased between 2005 and 2011 in both countries. For elective primary replacements most of the relative increase in Germany—11% (15 300 cases) overall—was caused by demographic factors, namely population aging.
After adjustment for demographics, a 3% increase caused by other factors remains. In the USA the total increase was 28%, significantly greater than in Germany. Demographic factors explain approximately half this increase.
Hip replacements for fracture increased by 15% in Germany due to demographic factors. However, all other determining factors resulted in a decrease of 8%. The net result of these two changes is the actual increase of 6% (2700 cases). In the USA too, non-demographic factors caused a drop in frequency of surgery nearly equal to the increase caused by demographic factors.
The greatest increase in hip replacements concerned revision replacements. In Germany these rose by 22% (4300 cases) overall. The influences of demographic and non-demographic causes were approximately equal. In the USA, revision hip replacements increased by 32% overall; adjustment for demographics leaves an increase of 18% (Table 2).
The number of primary knee replacement operations grew by 22% (27 000 cases) overall in Germany and 30% in the USA. These increases were caused by both demographic and non-demographic factors in both countries; non-demographic factors were slightly more significant in Germany.
The relative increase in knee revision replacements between 2005 and 2011 was 64% (6200 cases) in Germany and 50% in the USA. These sharp increases were mostly caused by non-demographic factors in both countries (Table 3).
Analyses that provide international comparisons are useful in ranking and assessing care in the context of differing health care systems. The USA was chosen as an example country with which to compare Germany for this article because appropriate data on individual patients was available, making it possible to perform a methodologically sound comparison.
The results of the research, which is based on independent analysis of individual patient data rather than evaluation of aggregated statistics prepared by others, show that hip replacements are performed more frequently in Germany than in the USA even after adjustment for differing demographics. However, the frequency of knee replacements was significantly lower in Germany than in the USA.
The number of operations performed increased during the observation period in both countries. For hip replacements, the overall increases were slightly smaller in Germany than in the USA. Most of the increase in elective primary replacements in Germany can be explained by demographic factors.
After adjustment for demographic factors, there was actually a decrease in primary replacements for fracture in older age groups. This may be related to changes in treatment strategies. The increase in revision replacements may have been a consequence of earlier increases in the primary replacement rate. Significantly higher increases that were unrelated to demographics were observed in the USA, where the baseline figures for elective primary replacements and revision replacements were lower, than in Germany.
Knee replacements also increased in Germany during the observation period, although there was a slight downward trend in 2010 and 2011. Non-demographic factors play a greater role in changes in knee replacement case numbers than for hip replacements. Revision knee replacements were performed approximately 1.6 times more frequently in 2011 than in 2005, independently of the effects of population aging; this can be seen as a consequence of earlier increases in primary replacements.
In the USA there were particularly sharp increases in total knee replacements from 2006 to 2008. These were caused by both demographic and non-demographic factors.
These results seem plausible when compared to other frequency figures reported on the basis of case numbers. Case numbers calculated for Germany using DRG statistics are comparable to those found in statutory quality assurance if differing definitions are taken into account (16).
There is also a good level of agreement with estimates based on health insurer data (17, 18), if entities defined in comparable ways are compared with each other. US publications, too, come to similar estimates of national frequencies for the USA (19–21).
Reliable knowledge can only be obtained from international comparisons if certain methodological requirements are met: in addition to taking into account countries’ differing demographics, clinical entities must be referred to appropriately, without overlooking the differing features of individual classification systems. Analyses can only be performed on the basis of representative, comparative individual patient data. Calculating numbers of procedure codes without using individual patient data can result in significant data distortion due to multiple counting of individual cases.
The USA was chosen as an example country with which to compare Germany for this extensive case-related analysis because appropriate data was available. It was not possible to provide comparisons based on individual patient data from other industrialized countries in this study. However, published rates for other countries have also been collated and are compared in eTables 4 and 5. The crude frequency of hip replacements in Germany is higher than in Sweden, Norway, the Netherlands, England and Wales, or Australia but lower than in Switzerland. For knee replacements, German rates are similar to those of Switzerland. However, lower rates are reported for England and Wales, the Netherlands, Denmark, and Norway, while higher rates are reported for Australia. Almost all data sources show that surgery numbers increase over time. Günther et al. (22) have also collated rates for various countries and conclude that Germany is in the top third of industrialized countries for both hip and knee replacements. However, it must be remembered that the extent to which crude figures can be interpreted is severely limited by demographic differences. Of all the countries examined, Germany has the highest proportion of inhabitants aged 65 or older (eTables 4 and 5).
The causes of the differences in hip and knee replacement frequency in Germany compared to the USA found in this research that are independent of demographics can only be a subject of speculation. It is possible that differences between health care systems affect access to joint replacement surgery. Because the USA has a higher proportion of uninsured individuals and significantly higher copayments for inpatient treatment, lower numbers of surgeries would be expected, as seen for hip replacements. This is not true of knee replacements, however. Differences in risk factor epidemiology may play a role here. For example, the proportion of overweight individuals, who are at greater risk of needing knee replacement, is higher in the USA than in Germany (23).
Changes in case numbers over time did not result from demographic factors alone in either country. Changes in numbers of elective primary replacements that are not caused by demographic factors may be the result of epidemiological factors. For example, in both Germany and the USA there is evidence of an increase in the prevalence of arthritis (24–26). However, it is also likely that surgery is being indicated more frequently as a result of medical and technical advances. In particular, the lower risks of surgery (e.g. thanks to less aggressive surgery and anesthesiology techniques) should be considered; these make it possible to provide such care even for patients with moderately increased risk. A change in demand by patients is also a possible cause, due to such surgery being seen as less risky, for example. Frequently discussed changes in supply by care providers, e.g. as a result of DRG introduction, may account for a further portion of changes unrelated to demographics but explain no more than some of the overall increase.
Due to the availability of suitable data, this article compares surgery frequencies in Germany and the USA only.
In order to rank these results better, further comparisons should be made with other industrialized countries, to the extent that individual patient data is available.
The non-demographic causes of differing surgery frequencies in the countries investigated cannot be analyzed more closely on the basis of the selected study design. In addition, it should be noted that there are differences between the German and US health care systems which must be borne in mind when interpreting the results.
In cross-sectional analysis there are considerable differences between the frequencies of both hip and knee replacements in Germany and the USA. While more people undergo hip replacement surgery in Germany, the frequency of knee replacements was significantly higher in the USA.
The number of operations performed increased in both countries during the observation period, from different baseline levels. In Germany, changes in numbers of primary hip replacements were mostly caused by demographic factors, while larger increases unrelated to demographics were observed in the USA. Non-demographic factors affected numbers of primary knee replacements to similar extents in both countries.
This analysis cannot determine whether too many or too few joint replacements are performed in Germany or the USA. Long-term studies must investigate the extent of any overtreatment, undertreatment, or incorrect treatment; such studies must measure medical benefit in terms of target parameters such as functional outcome or quality of life. In light of the results of this research, general statements that too many joint replacements are performed in Germany should be interpreted with care. Discussion of changes in case numbers over time must take greater account of the effect of demographic factors.
Conflict of interest statement
The Department of Structural Advancement and Quality Management in Health Care (Strukturentwicklung und Qualitätsmanagement im Gesundheitswesen) is sponsored by Helios-Kliniken GmbH.
Manuscript received on 11 December 2013, revised version accepted on 7 April 2014.
Translated from the original German by Caroline Devitt, M.A.
Prof. Dr. med. Thomas Mansky
Structural Advancement and Quality Management in Health Care
Technical University Berlin
10623 Berlin, Germany
@For eReferences please refer to:
eTables and eBox available at:
|1.||Hibbeler B: Krankenhäuser: „Wer Menge anreizt, kriegt Menge“. Dtsch Arztebl 2013; 110(43): A-2002/B-1770/C-1733. VOLLTEXT|
|2.||Flintrop J: Krankenhausfinanzierung: Wasser auf die Mühlen der Krankenkassen. Dtsch Arztebl 2013; 110(16): A-749/B-657/C-657. VOLLTEXT|
|3.||Mohrmann M, Koch V: Hohe Leistungsmengen – Direktverträge und Rechtehandel als Lösungen für den Krankenhausbereich. In: Klauber J, Geraedts M, Friedrich J, Wasem J: Krankenhaus-Report 2013. Mengendynamik: mehr Menge, mehr Nutzen? Stuttgart: Schattauer 2013. PubMed Central|
|4.||OECD: Health at a Glance 2011: OECD Indicators. OECD-Publishing 2011. http://dx.doi.org/10.1787/health_glance-2011-en (last accessed on 24 March 2014). CrossRef|
|5.||OECD: Health at a Glance 2013: OECD Indicators, OECD Publishing 2013. http://dx.doi.org/10.1787/health_glance-2013-en (last accessed on 24 March 2014). CrossRef|
|6.||OECD: Health at a Glance: Europe 2010, OECD Publishing 2010. http://dx.doi.org/10.1787/health_glance-2010-en (last accessed on 24 March 2014). CrossRef|
|7.||OECD: Health at a Glance: Europe 2012, OECD Publishing 2012. http://dx.doi.org/10.1787/9789264183896-en (last accessed on 24 March 2014). CrossRef|
|8.||Forschungsdatenzentren der statistischen Ämter des Bundes und der Länder: DRG-Statistik 2005 bis 2011. Wiesbaden: Forschungsdatenzentren der statistischen Ämter des Bundes und der Länder. www.forschungsdatenzentren.de/bestand/drg/index.asp (last accessed on 3 December 2012)|
|9.||Agency for Healthcare Research and Quality (AHRQ): Healthcare Cost and Utilization Project (HCUP). HCUP Nationwide Inpatient Sample (NIS). 2005–2011. Rockville, MD: Agency for Healthcare Research and Quality. www.hcup-us.ahrq.gov/nisoverview.jsp (last accessed on 3 December 2012).|
|10.||AQUA-Institut: Anwenderinformation QS-Filter (QS-Spezifikation 2013 SR 1). Göttingen: AQUA-Institut 2012. www.sqg.de/datenservice/spezifikationen-downloads/verfahrensjahr-2013/anwenderinformation-qs-filter-2013-sr-1.html (last accessed on 3 December 2012).|
|11.||Statistisches Bundesamt: Bevölkerung und Erwerbstätigkeit 2011. Bevölkerungsfortschreibung auf Grundlage der Volkszählung 1987 (Westen) bzw. 1990 (Osten). Fachserie 1, Reihe 1.3. Wiesbaden: Statistisches Bundesamt 2013.|
|12.||United States Census Bureau, Population Division: Table 1. Annual Estimates of the Resident Population by Sex and Five-Year Age Groups for the United States: April 1, 2000 to July 1, 2009 (NC-EST2009–01). Washington, DC: US Census Bureau 2010. www.census.gov/popest/data/historical/2000s/vintage_2009/index.html (last accessed on 5 September 2013).|
|13.||United States Census Bureau, Population Division: Table 1. Annual Estimates of the Resident Population by Sex and Five-Year Age Groups for the United States: April 1, 2010 to July 1, 2011 (NC-EST2011–01). Washington, DC: US Census Bureau 2012. www.census.gov/popest/data/historical/2010s/vintage_2011/index.html (last accessed on 5 September 2013).|
|14.||Friedrich J, Günster C: Determinanten der Casemixentwicklung in Deutschland während der Einführung von DRGs (2002 bis 2004). In: Klauber J, Robra B-P, Schellschmidt H, eds.: Krankenhausreport 2005. Schwerpunkt: Wege zur Integration. Stuttgart: Schattauer 2006.|
|15.||Nowossadeck E: Population aging and hospitalization for chronic disease in Germany. Dtsch Arztebl Int 2012; 109: 151–7. VOLLTEXT|
|16.||AQUA-Institut. Qualitätsreport 2011: Göttingen: AQUA – Institut für angewandte Qualitätsförderung und Forschung im Gesundheitswesen 2012.|
|17.||Bitzer EM, Grobe TG, Dörning H, Schwartz FW: BARMER GEK Report Krankenhaus 2010. Schwäbisch Gmünd: BARMER GEK 2010.|
|18.||Schäfer T, Pritzkuleit R, Jeszenszky C, Malzahn J, Maier W, Günther KP, Niethard F: Trends and geographical variation of primary hip and knee joint replacement in Germany. Osteoarthritis Cartilage 2013; 21: 279–88. CrossRef MEDLINE|
|19.||Losina E, Thornhill TS, Rome BN, Wright J, Katz JN: The Dramatic increase in total knee replacement utilization rates in the United States cannot be fully explained by growth in population size and the obesity epidemic. J Bone Joint Surg Am 2012; 94: 201–7. CrossRef MEDLINE PubMed Central|
|20.||Kim S: Changes in surgical loads and economic burden of hip and knee replacements in the US: 1997–2004. Arthritis Rheum 2008; 59: 481–8. CrossRef MEDLINE|
|21.||Zhan C, Kaczmarek R, Loyo-Berrios N, Sangl J, Bright RA: Incidence and short-term outcomes of primary and revision hip replacement in the United States. J Bone Joint Surg Am 2007; 89: 526–33. CrossRef MEDLINE|
|22.||Günther KP, Jeszenszky C, Schäfer T, Hannemann F, Niethard F: Hüft- und Kniegelenkersatz in Deutschland – Mythen und Fakten zur Operationshäufigkeit. Das Krankenhaus 2013; 9: 927–33.|
|23.||Gellner R, Domschke W: Epidemiologie der Adipositas. Chirurg 2008; 79: 807–10. CrossRef MEDLINE|
|24.||Robert Koch-Institut (eds.): Daten und Fakten: Ergebnisse der Studie „Gesundheit in Deutschland aktuell 2010“. Beiträge zur Gesundheitsberichterstattung des Bundes. Berlin: Robert-Koch-Institut 2012.|
|25.||Nguyen US, Zhang Y, Zhu Y, Niu J, Zhang B, Felson DT: Increasing prevalence of knee pain and symptomatic knee osteoarthritis: survey and cohort data. Ann Intern Med 2011; 155: 725–32. CrossRef MEDLINE|
|26.||Suri P, Morgenroth DC, Hunter DJ: Epidemiology of osteoarthritis and associated comorbidities. PM&R 2012; 4: 10–9. CrossRef MEDLINE|