The Swedish Hip Arthroplasty Register

Johan Krrholm, M.D. PhD
Professor in Orthopaedics,
Head of the Swedish Hip Arthroplasty Register

Gran Garellick, M.D., PhD
Peter Herberts, M.D., PhD
Department of Orthopaedics,
Sahlgrenska University Hospital

In 1977 a pilot study was initiated to record all primary and revision hip arthroplasties in Swedish hospitals. Experiences from this study resulted in the Swedish Hip Arthroplasty Register founded in 1979. From the very beginning, all individual revision arthroplasties were identified with a social security number. Demographic data, details about surgical technique, implant used, operating unit and the type and history of the previously used implant or implants were recorded. Until 1992 data about primary hip arthroplasties were recorded on a hospital basis. This meant that it was not possible to connect each individual implant to a social security number which resulted in some uncertainties in the outcome analyses. Implant survival was calculated by use of information from other governmental register and statistically based models1. In a thorough evaluation it did; however, turn out that the recorded results were valid6. From 1992, all primary total hips were also recorded in more detail. This meant that each surgical procedure could be connected to patient demographics, type of incision, a specific implant and technique of fixation. In 1999 the recording was further expanded by including more details about the implants. Information about sizes, offset, and implant materials became available. These registers cover the entire country i.e. all public and private hospitals in Sweden participate.

In 2002 patient-related outcome was added4. Patients scheduled for total hip arthroplasty filled in the EQ-5D form as a screening of health-related quality of life (HRQoL). In addition, pain and overall satisfaction was indicated on a VAS-scale. This routine now covers 74 of the total 78 units (95%). In 2005 a further step was taken when all hemiarthroplasties were recorded. Use of the same organization as employed for total hips resulted in an immediate nationwide coverage.

At present, most units report their primary data on line with use of a specific web application. Re-operations (any kind of further surgical procedure after the primary operation) and revisions (any new surgical procedure where the entire implant or parts of it are exchanged or removed) are reported. A case record is sent to a central unit for detailed analysis.

The report contains a general section including demographic data, frequency of procedures, choice of implant and fixation, and survival of the different implants or surgical techniques on a national basis (Figures 1 and 2). The same data are also reported regionally and to the individual units. During the past years we have also included a number of specific in-depth analyses. Examples of such analyses are periprosthetic fracture3, dislocations, international comparisons5, factors influencing outcome in young patients (up to 50 years of age) and influence of stem size on revision rates in the three most commonly used cemented stems2.

Figure 1. Survival based on revision due to aseptic loosening of cemented THR. Hips with primary osteoarthritis operated 1992 and later have about 6% better 15-year survival than those operated earlier.


Figure 2: Survival based on revision due to aseptic loosening of uncemented THR in hips with primary osteoarthritis. Overall and without adjustment for other confounders than diagnosis, uncemented fixation has not performed as well as cemented. There is, however an improvement in the cohort operated 1992 and later.

When different operating units are compared, it is important to define the patient population with regard to expected outcome. In Sweden more difficult cases are admitted to hospitals with specific expertise or resources: usually University, but also County hospitals. We have computed a simple case-mix factor, which should be used to adjust the expected outcome and to estimate the amount of resources needed. Based on a number of regression analyses, we have defined a simple case-mix indicator related to risk of revision. We found that females with primary osteoarthritis, at an age of 60 years or older had the best outcome. In this group, the risk of revision was reduced to about 50 % when compared with all other patients regardless if one used revision at 2, 5 or 10 years as the outcome variable. The observed number of revisions increased stepwise depending on gender, age and type of osteoarthritis up to fourfold as one proceeded from the most favourable situation (female with primary OA 60 years of age or more) to the most unfavourable (male, secondary osteoarthritis, younger than 60 years of age). This case-mix indicator is reported each year for all participating hospitals (Table 1).

Table 1. Case-mix indicator based on gender, age and diagnosis. Observed revision rate due to aseptic loosening/osteolysis at two, five and ten years if none, one, two or three positive factors (female, age >= 60 years, primary osteoarthritis) are present. Relative risks of revision (compared to the remaining cases) if all positive factors are present at bottom. Analysis based on 169 622 THR performed 1992-2006.


Observed revision rate (%)


0-2 yrs

0-5 yrs

0-10 yrs

Positive factors n


















Risk reduction SE:

female, POA, >=60 yrs

0.53 0.05

0.52 0.05

0.50 0.03

Another important function of the hip register has been the publication of indicators, which are related to the quality of the outcome. Three such important indicators are: 1) number of reoperations (any kind of surgical intervention) within two years after the primary operation, 2) implant survival at ten years and 3) life quality gained at the one year follow-up according to EQ-5D. In the latest report, the openly reported result indicators have increased to eight. These data are open to the public and are published for each surgical unit. From this evaluation, it has become evident that some clinics have an unexplained high frequency of reoperations because of early dislocations, some have more infected cases and in other units, the patient satisfaction does not correspond to the expected outcome. Such information has proved to have a positive influence on the quality of the health care. This information has initiated a thorough effort to improve performance. We think that this open comparison is important to continuously improve the outcome of total hip arthroplasty on a nationwide basis. In Sweden there is a general agreement, also within the orthopaedic profession, that the use of the recorded data is cost-effective and should be continued.

To improve further analysis, we intend to expand the recording of primary data in 2008 (BMI, ASA, waiting time). We have also initiated a Nordic collaboration (Nordic Arthroplasty Register Association) between the different national hip registers in order to perform more comprehensive analyses and to allow us to make conclusions with regard to infrequently used implants and procedures.


  1. Herberts P., Malchau H. How outcome studies have changed total hip arthroplasty practices in Sweden. Clin Orthop Relat Res. 1997 Nov;(344):44-60
  2. Krrholm J., Garellick G., Lindahl H., Herberts P. Improved analyses in the Swedish Hip Arthroplasty Register. Vetenskaplig utstllning p 74th Annual Meeting of the American Academy of Orthopaedic Surgeons, San Diego, USA, March 14-18, 2007
  3. Lindahl H., Malchau H., Herberts P., Garellick G. Periprosthetic femoral fractures classification and demographics of 1049 periprosthetic femoral fractures from the Swedish National Hip Arthroplasty Register. J Arthroplasty. 2005 Oct;20(7):857-65.
  4. Malchau H., Garellick G., Eisler T., Krrholm J., Herberts P. Presidential guest speaker: the Swedish Hip Regsitry: Increasing the sensitivity by patient outcome data. Clin Orthop 2005;441:19-29
  5. Ostendorf M., Johnell O., Malchau H., Dhert W.J.A., Schrijvers A.J.P., Verbout A.J. The epidemiology of total hip replacement in The Netherlands and Sweden: present status and future needs. Acta Orthop Scand 2002;73(3):282-286
  6. Sderman P., Malchau H, Herberts P, Johnell O. Are the findings in the Swedish National Total Hip Arthroplasty Register valid? A comparison between the Swedish THA register, the National Discharge Register and the National Death Register. J Arthroplasty 2000;15:884-889

Submit Community Content

If you have orthopedic information that you would like to share with the Orthogate Community, please register/login and submit your news, event, job, article, case or workshop from the Submit Content menu under the My Account area. Learn more!