Trends and Economic Impact of Hip and Knee Arthroplasty in Central Europe: Findings from the Austrian National Database

Arthroplasty registers were originally established in Scandinavia to receive clinically relevant information from pooled data, to improve quality and reduce revision surgeries, with socioeconomic benefit. In Austria, where the highest rate of total knee arthroplasties (TKA) per inhabitant of all OECD countries was reached in 2014, arthroplasties are centrally reported since 2009. Study purpose was to perform the first analysis of the Austrian database, aiming to obtain data on trends in arthroplasty in Austria over time in relation to demographic development. Between 2009 and 2015 an almost continuous increase of total hip arthroplasties (THA; 18.052) by 14% and TKA (17.324) by 13% were observed, representing 210 THA and 202 TKA per 100k inhabitants in 2015. A similar increase was found for revision surgeries, with 1.290 re-implanted THA (7.1% of all THA) and 919 re-implanted TKA (5.3% of all TKA) in 2015. Implantation of mega or tumor prosthesis for the knee and hip joint remained constant and was mainly performed in two university hospitals. Patellar resurfacing decreased by 31.6%. Demographic development will further increase the number of primary and revision surgeries. Inclusion of more detailed information on used and revised components was established and will improve efficacy in quality control.


Methods
Number of joint replacements and revision surgeries of the hip (primary arthroplasty, tumor/resection replacement, explantation, cement spacer implantation/explantation, re-implantation, exchange of not bone-anchored implant components) and knee joint (primary arthroplasty, tumor/resection replacement, explantation, cement spacer implantation/explantation, re-implantation, exchange of not bone-anchored implant components, patellar surface replacement) were centrally registered by the Austrian health authorities with respect to the performance-based hospital financing system 5 . The number is based on service codes transmitted by healthcare providers to receive payment for the provided medical services. Total number, sex, age group in 5 years steps, and the province were recorded. Descriptive analysis of the database as well as subgroup analysis concerning patients´ age, sex, and chronological sequence was performed. Extraction of implantation data for singular hospitals was not possible from the data. Comparison to data on demographic trends in Austria, which are published yearly by "Statistik Austria" (formerly, Austrian Statistical Office) was performed 6 .
Statistical methods. Descriptive statistics were calculated from data delivered by the national health authorities and "Statistik Austria" institution. The calculated mean age was estimated assigning the median age for the age group to all patients belonging to an age group. IBM SPSS Statistics 20 was used for data analysis.
Ethics statement. No Institutional Review Board approval was required for this project.  Table 1). This represents 210 primary THA and 202 primary TKA per 100.000 inhabitants in 2015. Ninety-two percent of THA and 97% of TKA were performed in the age group 50-90 between 2009 and 2015 (Suppl. Figure 1). Between 2009 and 2015 the population aged 50-90 increased only by 1.1% to 3.3 million people, the population older than 65 years in Austria increased by 9.3%, and the population older than 70 years increased the strongest by 19.8% 6 Figure 3A). Implantation was mainly performed in two regions with  Figure 3B). Two implantation peaks were found for mega prosthesis and tumor prosthesis for knee and hip, the first around 15-20 years and the second in the population 60 years and older (Suppl. Figure 3C). The only implant type that clearly decreased in case numbers during the observation period was patellar resurfacing with a decrease of 31.6% between 2009 (n = 1983) and 2015 (n = 1356) (Suppl. Fig. 4).

Discussion
This manuscript provides the first analysis of a detailed report on the Austrian National database on knee and hip arthroplasty, in comparison with demographic development in Austria from 2009 to 2015.
We explored an increase in almost all implant-related surgeries of the knee and hip joint in Austria during the observation time from 2009 to 2015. Corresponding increases can be found in similar studies from most developed countries (Fig. 1) 7,8 . The increased number of implanted THA by 14% and the increased number of TKA by 13% over this period represents 210 THA and 202 TKA in 100k inhabitants in 2015. This number is relatively high in comparison to numbers from other countries previously published by our study group 9 . Demographic change, especially increasing proportion of elderly people, might be a main reason for this increase in operations. Studies in Germany, a neighboring country to Austria with a similar demographic change, have shown that the increase in age leads to an increase in multimorbidity and especially to an increase in the prevalence of osteoporosis which triggers musculoskeletal diseases [10][11][12] . On the other hand, it has been shown that technical innovation led to reduced surgical risk, resulting in lower threshold for indicating arthroplasties, and thus enabling an increased number of surgeries especially in elderly patients 13 . Our data revealed that the increment of implanted THA and TKA is disproportional compared to increase in Austria's population during the observation time (Fig. 3A). The OECD recently published, that Austria had the highest rate of TKA per inhabitant of all OECD countries in 2014 4 (Fig. 3B).  As stated before, demographic change alone can only represent part of explanation for increasing number of implanted prosthesis 4 . In part, ongoing development of medical care also seems to be responsible for increased intervention numbers. On the other hand, environmental development may lead to economic growth pressure, resulting in higher implant numbers than explored for comparable neighboring countries (Fig. 3B). Such a development should be avoided in a health system since it may result in 'overtreatment' with arthroplasties.
Mean age in Austria's population increased, whilst mean age of patients undergoing primary THA and TKA remained on steady state during the observation period, suggesting a slight trend towards younger arthroplasty receivers. A similar phenomenon was explored in the United States and Canada in earlier studies 8,14 .
Especially the disproportionately high increase in revision surgeries after THA and TKA (Fig. 2) seems to be a result of the combination of trends addressed above, higher implantation numbers, younger receivers, and increasing life expectancy, altogether resulting in an increased population at risk. Earlier authors have raised concern on survivorship of primary TKA and revision surgeries in younger patients, for which higher risk of early periprosthetic joint infection and aseptic mechanical failure after primary implantation as well as higher rates of aseptic failure after revision have been described 15,16 .
Implantations of mega prosthesis or tumor prosthesis for the knee and hip are usually performed in tertiary centers, where the underlying disease can be treated 17 , resulting in concentration of implant numbers in Vienna and Styria, where skeletal oncology centers are located (Suppl. Figure 3B). This data interpretation is supported by the two age peaks for both, implanted mega prosthesis or tumor prosthesis for the knee and hip, congruent to age peaks for incidence of osteosarcoma and Ewing sarcoma in young patients and chondrosarcomas and bone metastases in the elderly (Suppl. Figure 3C) [18][19][20] .
The Austrian database reveals a continuous decrease of 31.6% in the number of patellar resurfacing performed during the observation time. A similar trend has been described by our group in a meta-analysis of arthroplasty registers, as, for example in Sweden, whereas TKA is nowadays mostly performed without patella buttons 9 . The benefit of patellar resurfacing is discussed controversially in the scientific society; whilst the majority of surgeons (>90%) routinely resurface the patella in North America, most Austrian surgeons seem to only selectively resurface based upon patient factors or during revision surgeries 21 .
Study Limitations. The Austrian arthroplasty registry is based on data which are transmitted mainly for deposit with the public health system. Although this implicates high completeness of our dataset, it has lower granularity than conventional prosthesis registers on the patient-level as provided by Sweden or Germany. Although, networking back on individual cases (for inspection purposes) is possible, the data do not contain additional information except for the parameters presented. Therefore, no information is available on patients clinical condition, how surgery was performed, implant type and manufacturer, patient reported outcome scores, or follow up, (etc.) presenting a major limitation for more detailed investigations. However, it is possible to screen for higher revision or complication rates in order to perform an on-site peer review audit organized by the Austrian ministry of health.
In 2015, the service codes for revised knee or hip arthroplasty were further specified on the component of the prosthesis, which was revised. For comparability with data from the years before, the numbers for new codes were summed up. Notably there was a peak in number of THA re-implantations in 2015 ( Fig. 2A), which could result from transmission of service codes for each component when the complete system was changed. This would represent a reporting bias. This effect was not found for TKA re-implantations in 2015.
The authors believe that the implantation number for mega prosthesis reported to the Austrian health authorities is too high. This could represent a reporting bias, resulting from transmission of service codes for mega prosthesis when large revision surgeries of TKA or THA were performed.

Conclusion
Many parallels to earlier published databases and register results can be drawn from this Austrian National database analysis, and underline the validity of our data. A common consensus could be that numbers of arthroplasties are still increasing in developed countries; the peak in implantation numbers for primary implantations and revision surgeries after arthroplasty seems not to be reached yet. Knowledge on numbers of surgeries in this field is important for quality control in the Austrian public health system. Inclusion of more detailed information on used and revised components, as it is established since 2015, will improve efficacy in quality control.