Total knee arthroplasty (TKA) is a commonly used treatment for severe primary knee osteoarthritis (OA) that is refractory to conservative treatment. Despite the presumed increase in the use of this treatment modality in younger patients, there are few published data concerning this. The aim of our study was to examine the changes in the age- and sex-standardized incidence of TKA and unicondylar knee arthroplasty (UKA) in Finland during 1980–2006 and to identify factors that might affect the incidence during this period.
We obtained data on UKAs and TKAs from the Finnish Arthroplasty Registry and population data from Statistics Finland to analyze the incidence of UKAs and TKAs in Finland for the period 1980–2006. The effects of sex, age group, and hospital volume on the incidence of these procedures were also evaluated.
The annual cumulative incidence of UKAs and TKAs has increased rapidly from 1980 to 2006 among 30–59-year-old inhabitants of Finland. For UKAs, the incidence increased from 0.2 per 100,000 inhabitants to 10 per 100,000, and for TKAs, the incidence increased from 0.5 per 100,000 to 65 per 100,000. The incidence remained higher among women during the entire study period. Most of the increase occurred among patients ages 50–59 years. The incidence grew more rapidly in low-volume and intermediate-volume hospitals.
We demonstrated a rapid increase in the incidence of arthroplasty among patients with primary knee OA in Finland, especially in those ages 50–59 years. There was no single explanatory factor behind this finding, although some of the growth might be due to the increased incidence noted in low- and intermediate-volume hospitals.
Total knee arthroplasty (TKA) is a treatment method commonly used in patients with severe osteoarthritis (OA) that is refractory to conservative treatment. Good long-term results with cemented condylar prostheses have been described (1–3). The majority of patients in these studies, however, have been older than 60 years. Meanwhile, the so-called baby-boom generation (the generation born after World War II) is growing older, placing increasing demands on the health care system (4).
The estimated demand for primary TKA has been predicted to grow by 673% to 3.48 million procedures in the US by the year 2030. Similar projections have also been proposed in other Western countries (5–9). Wells et al (8) reported an increase in the incidence of TKA in younger patients (45–64 years of age) with primary OA in Australia between 1988 and 1998. Those investigators also showed a geographic variation in the incidence of TKA between the Australian states and territories. Jain et al (5) reported that the incidence of TKA had increased 54% in the group ages 50–59 years and 95% in the group ages 40–49 years in the US between the 1990–1993 cohort and the 1998–2000 cohort. They also noted that the increase in incidence was more rapid in high-volume centers (≥400 total knee replacements per year) than in low-volume centers. Similar data on the incidence of unicondylar knee arthroplasty (UKA) in young patients have not been published.
In the present study, we analyzed changes in the age group– and sex-standardized incidence of UKA and TKA in Finland between the years 1980 and 2006. In addition, we sought to identify the factors that might have effects on the incidence during the study period.
PATIENTS AND METHODS
The Finnish population and health care system.
During the study period, the total population of Finland increased from 4,787,778 inhabitants in 1980 to 5,276,955 inhabitants in 2008. The Finnish health care system is funded through general taxation resources and is provided by the municipalities. There is considerable emphasis on primary care, preventive services, and documentation (10).
The Finnish Arthroplasty Register.
Since 1980, the Finnish Arthroplasty Register has been collecting information on total knee replacements. Consequently, health care authorities, institutions, and orthopedic units are obliged by law to provide the information that is essential for the maintenance of their registry. The coverage of the Finnish Arthroplasty Register was estimated for the period 1994–1995 by comparing its data with those of the discharge registries of the participating hospitals, and was found to cover 90% of implantations. Since 1995, the data from the registry have been compared with those from hospital discharge registries every few years. Currently, 98% of implantations are recorded (refs.11–13, and M. Peltola: personal communication [National Research and Development Centre for Welfare and Health in Finland; July 27, 2008]).
Description of the population data.
Every inhabitant of Finland has a unique social security number. This number indicates that its holder is entitled to the public health care services of Finland. Statistics Finland keeps an accurate, up-to-date civil registry of the inhabitants of Finland. In practice, there are no known significant numbers of illegal immigrants in Finland. Therefore, data derived from such a source are unlikely to be a confounding or biasing factor in our study.
All TKAs and UKAs performed on patients under the age of 60 years because of primary knee OA that had been entered in the Finnish Arthroplasty Registry between 1980 and 2006 were included in the present study. All other diagnoses were excluded. Population data for Finland during the period 1980–2006 were obtained from Statistics Finland. All of the population data, including the population data for each municipality and each county, were divided into cohorts by age (30–39 years, 40–49 years, and 50–59 years) and were subgrouped according to sex.
Analysis of the incidence of knee arthroplasty.
Annual incidence rates for TKAs and UKAs were calculated. Bilateral arthroplasties were handled as 2 separate operations in each analysis. The general incidence (in subjects ranging in age from 30 years to 59 years), the incidence by sex, and the incidence by age cohort for the knee arthroplasties were calculated from the number of operations performed in the entire study population for both sexes and for each of the 3 age cohorts. The general incidence was presented as the number of operations performed per 100,000 inhabitants ages 30–59 years. Incidence by sex was presented as the number of operations performed per 100,000 men or women ages 30–59 years. The incidence by age groups was presented as the number of operations performed per 100,000 inhabitants ages 30–39 years, 40–49 years, and 50–59 years, respectively.
We also calculated the incidence for each public hospital. All inhabitants of Finland have to be registered as a resident of some county or municipality. Each public hospital has its own catchment area (consisting of 1 or more counties and/or municipalities), and all patients from those counties and/or municipalities are primarily referred to the specific hospital in whose administrative area they reside. In addition, there are 2 major foundation-based private hospitals that treat public patients in Finland. In order to apportion the correct incidence in the respective catchment area for each communal hospital, TKAs and UKAs performed in the private institutions were added to the patients' “own” communal hospitals operation numbers, based on the municipality or county in which they lived. Thus, the incidence of TKAs and UKAs in each catchment area was calculated according to the number of arthroplasties performed in both the respective community hospitals and the 2 foundation-based hospitals for each particular catchment area.
Analysis of the effect of hospital volume on incidence.
For this analysis, the study period was restricted to the years 2000 through 2006 because only in the twenty-first century did the hospitals start to demonstrate definite differences between annual numbers of knee arthroplasties. Hospitals were divided into low-, intermediate-, and high-volume centers according to the number of TKAs performed among all hospitals in Finland during 2006. All knee arthroplasties, regardless of the age group or the indication for the operation, were taken into account when dividing the hospitals into these 3 categories. Hospitals that performed <100 knee replacements, 100–400 knee replacements, and >400 knee replacements during 2006, regardless of the ages of the patients, were categorized as low-, intermediate-, and high-volume centers, respectively. All hospitals belonged to the same volume category throughout the entire study period as was assigned according to their volume in 2006. All university hospitals except 1 belonged to the high-volume center category, and almost every central hospital belonged to the intermediate-volume center category. The low-volume center category mainly consisted of smaller rural hospitals.
We first described the data using descriptive statistics. We assessed trends in the general incidence of TKAs and UKAs between the years 1980 and 2006. Then, we examined the incidence of TKAs and UKAs separately for sex, hospital volume, and age groups. Finally, we built multivariable Poisson regression models to assess sex, hospital volume, and age groups with regard to the incidence of TKAs and UKAs. The results are expressed as incidence rate ratios (IRRs) with 95% confidence intervals (95% CIs). P values less than 0.05 were considered statistically significant. The statistical analyses were conducted with the SPSS 16.0 software and Stata 8.2 software (StataCorp).
Between the years 1980 and 2006, a total of 95,577 primary knee arthroplasties were performed in Finland, and primary OA was an indication in 74,715 of these operations (78.2%). Among these operations, 7,829 TKAs and 1,132 UKAs (total of 8,961 knee arthroplasties) were performed for primary OA in patients under the age of 60 years. Female patients represented 5,797 of these cases (64.5%). The mean age of the patients was 54.9 years (range 30–59 years) in the TKA group and 53.9 years (range 30–59 years) in the UKA group (Table 1).
Table 1. Knee arthroplasties performed in Finland during the study period*
OA = osteoarthritis.
Total number of primary knee arthroplasties
No. (%) with indication of primary OA
No. (%) of patients <60 years of age
Total knee arthroplasty
Unicondylar knee arthroplasty
Total knee arthroplasty
Unicondylar knee arthroplasty
Mean age of the patients (range), years
Total knee arthroplasty
Unicondylar knee arthroplasty
General incidence of knee arthroplasty.
There was a 130-fold increase in the incidence of TKAs over the entire study period, from 0.5 operations per 100,000 inhabitants (ages 30–59 years) to 65 per 100,000 (Figure 1). The incidence increased very rapidly between 2001 and 2006, from 18 per 100,000 to 65 per 100,000. The IRR for the annual increase in the general incidence of TKAs was 1.19 (95% CI 1.19–1.20; P < 0.001).
The incidence of UKAs increased from 0.2 operations per 100,000 inhabitants to 10 per 100,000 (Figure 1). The increase in the number of UKAs was less than that observed for the number of TKAs. There was, however, a period of rather fast growth toward the end of the study period (from 2000 to 2006) (Figure 1). The IRR for the annual increase in the general incidence of UKAs was 1.26 (95% CI 1.24–1.28; P < 0.001). The incidence of TKAs was higher than the incidence of UKAs, with an IRR of 6.92 (95% CI 6.50–7.36; P < 0.001).
Incidence by sex.
TKAs were performed more frequently in women than in men (Figure 2). The incidence of TKAs increased among both sexes, but it remained consistently higher among women throughout the study period, and there was a significant interaction between sex and time with regard to TKAs (P < 0.001). During the last 10 years of the study period, the incidence was 1.6–2.4-fold higher in women than in men.
UKAs were also performed more often in women than in men (Figure 2). Until year 1999, there were no major differences between the sexes. After 2000, however, there was a larger increase in the incidence of UKAs among women than among men, but there was still no significant interaction between sex and time with regard to UKAs (P = 0.036).
Incidence by age group.
In patients ages 30–39 years, there was no major increase in the incidence of TKAs over the study period, with an IRR of 1.09 (95% CI 1.04–1.14; P < 0.001) (Figure 3). In patients ages 40–49 years, the increase was also quite limited, ranging from 0 per 100,000 inhabitants to 18 per 100,000, and the IRR was 1.19 (95% CI 1.18–1.21; P < 0.001). In patients ages 50–59 years, however, there was a large increase in incidence during the study period, from 1.5 per 100,000 in 1980 to 160 per 100,000 in 2006, and the IRR was 16.8 (95% CI 12.2–23.2; P < 0.001). There was a significant interaction between the age groups and time with regard to TKAs (P < 0.001).
In patients who underwent UKA, there was a similar variation in incidence among the age groups as was observed in those who underwent TKA (Figure 3). The incidence of UKAs remained low in patients ages 30–39 years (varying from 0 per 100,000 to 0.3 per 100,000), and the IRR was 1.30 (95% CI 1.06–1.60; P < 0.001). In patients ages 40–49 years, the incidence grew slightly over the period from 2000 to 2006 (from 0 per 100,000 to 5 per 100,000), with an IRR of 1.33 (95% CI 1.28–1.40; P < 0.001). However, the incidence grew most rapidly in the oldest age group. In this group, the incidence increased from 0.5 per 100,000 to 24 per 100,000, and the majority of the growth occurred during the period 2000–2006. The IRR in this age group was 1.23 (95% CI 1.21–1.24; P < 0.001). Comparing age group 40–49 years with age group 30–39 years, the IRR for an increase in the incidence of UKA was 24.4 (95% CI 10.8–55.2; P < 0.001). Comparing age group 50–59 years with age group 30–39 years, the IRR was 147.2 (95% CI 66.0–328.5; P < 0.001). There was no significant interaction between the age groups and time with regard to UKAs (P = 0.822).
Incidence by hospital volume.
The incidence of TKA grew in every group during the period 2000 to 2006. The highest increase was seen in the low-volume and intermediate-volume hospitals. Between the intermediate- and high-volume centers, the IRR was 1.23 (95% CI 1.16–1.31; P < 0.001), and between the low-volume and high-volume centers, the IRR was 1.23 (95% CI 1.13–1.34; P < 0.001) (Figure 4). There was no significant interaction between the hospital groups and time with regard to TKAs (P = 0.071).
The incidence of UKA varied between 0 to 13 operations per 100,000 during the study period. The increase was highest in the low-volume hospital group from the year 2004 and there were significant differences in the incidence between the hospital groups and time with regard to UKAs (P < 0.001).
We found that the incidence of TKAs and UKAs in Finland increased very rapidly in patients with primary knee OA who were under the age of 60 years during the study period 1980–2006. In addition, the incidence of TKA grew faster than that of UKA. Regardless of the arthroplasty method, the incidence was consistently higher in women than in men throughout the entire study period. Most of the increase in incidence occurred in patients ages 50–59 years. The incidence of TKAs increased more in the intermediate-volume and low-volume centers than in the high-volume centers during the 2000–2006 study period.
Registry-based studies have certain limitations. It has been shown that the Finnish Arthroplasty Register covers 90% of hip implants for the year 1995, and after that, the coverage increased to 98% (13). This might explain a minor part of the increase shown in our study. Because of the accurate civil register of Statistics Finland and the low level of illegal immigration in Finland, the population data can be assumed to closely correspond to the actual population. However, divergence between international health care systems might confound direct comparisons between reports from different countries. The boundaries between low-, intermediate-, and high-volume centers were arbitrary, but they adapt the previous joint registry–based studies done in Finland and in the US (5, 14), and they formed well-distinguished groups of hospitals performing arthroplasties during the twenty-first century.
There have been several reports on the general incidence of TKA in the restricted population, but most of them focus on elderly patients with primary OA of the knee (5–8). In some of those studies, the incidence of knee arthroplasty has been shown to have increased by up to 10% per year for the last 25 years (15).
In our study, a rapid increase in the incidence of TKAs was determined among younger patients in Finland, especially over the last 10 years. Jain et al (5) demonstrated the same phenomenon, although their study period ended in 2000. In an analysis from a community joint registry in St. Paul, Minnesota, Gioe et al (16) reported that over time, the total percentage of young patients (under 55 years of age) in the overall registry TKA population increased from 4.6% in 1991 to 17.1% in 2005. Their study design was similar to ours. In our study, growth was especially fast between 2000 and 2006. It must also be noted that data collected from nationwide registers describes the true incidence of UKA and TKA in different age groups more accurately than do data from narrower regional studies.
In a recent population-based study, the incidence of OA especially among women was found to have decreased in Finland during the period 1980–2000 (17). Despite the decreasing incidence of knee OA in Finland, a reciprocal rapid increase in TKAs was found in the present study. Possible explanations for this phenomenon include the high functional and quality-of-life demands of younger patients under the age of 60 years. Another reason could be that the baby boomers may opt to undergo elective operations at an earlier stage, when they have milder symptoms, than previous generations. The baby-boom generation is well-educated and aware of the contemporary treatment options, which include TKA and UKA. Nowadays, information about the different treatment options is readily available via the internet and other media. Good long-term results in terms of implant survival in TKAs among older age groups, the continuous development of anesthetic and surgical techniques, in addition to the constant updating and retraining of orthopedic surgeons may all have led to the observed increase in the incidence of knee arthroplasty.
Simultaneous with the rapid increase in the incidence of knee arthroplasties, we witnessed a clear decrease in the incidence of high tibial osteotomy in Finland (18). This may partly explain the growth in incidence in both arthroplasty groups.
In our study, there was a higher age group–standardized incidence of knee arthroplasties in women than in men. Similar findings have recently been published regarding TKAs (5–7). We found that the actual growth in the incidence of knee arthroplasties among younger patients took place in the subgroup of patients ages 50–59 years. There was, however, a minor increase in the incidence of UKAs among patients ages 40–49 years. Nonetheless, the incidence of total knee arthroplasty in the same group remained at the same low level throughout the study period. This is consistent with the results reported recently by Jain et al (5). Our study population consisted of only patients with primary knee OA as the recorded indication for arthroplasty. If we had included data for patients with secondary (posttraumatic) OA, rheumatoid arthritis, and osteonecrosis, the incidence of knee arthroplasties would have been higher in the youngest age groups.
In our study, the most rapid growth in the incidence of knee arthroplasties occurred in the low-volume and intermediate-volume hospitals. In contrast to our findings, Jain et al (5) reported that the increase in incidence was fastest in the high-volume hospitals in the US. One could reason that smaller hospitals with the combination of fewer peer review indications for knee arthroplasty on the one hand and less limited local resources on the other would be expected to vary more than larger-volume hospitals.
In conclusion, the age group– and sex-standardized incidence of TKAs and UKAs has grown rapidly in patients in Finland with primary OA who are younger than 60 years, and most of the increase took place in the 50–59-year age group. This phenomenon has been especially strong during the twenty-first century. There is no single explanatory factor for this growth. Some of the increased incidence can be explained by hospital volume. Long-term results in young patients may differ from those reported in older patients, and the risk for revision surgery may be higher. More long-term data are warranted before the widespread use of TKA in younger patients can be justified.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Leskinen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Leskinen, Eskelinen, Huhtala, Paavolainen, Remes.
Acquisition of data. Leskinen, Eskelinen, Huhtala, Paavolainen, Remes.
Analysis and interpretation of data. Leskinen, Eskelinen, Huhtala, Paavolainen, Remes.