You are here

Article Text

Article menu
Clinical and epidemiological research
Extended report
Cardiac and thromboembolic complications and mortality in patients undergoing total hip and total knee arthroplasty
  1. Jasvinder A Singh1,2,3,4,
  2. Matthew R Jensen1,
  3. William Scott Harmsen1,
  4. Sherine E Gabriel1,
  5. David G Lewallen2
  1. 1Department of Health Sciences Research, Mayo Clinic School of Medicine, Rochester, Minnesota, USA
  2. 2Department of Orthopedic Surgery, Mayo Clinic School of Medicine, Rochester, Minnesota, USA
  3. 3Rheumatology Section, Medicine Service, VA Medical Center, Birmingham, Alabama, USA
  4. 4Division of Rheumatology, Department of Medicine, University of Alabama, Birmingham, Alabama, USA
  1. Correspondence to Jasvinder A Singh, Division of Rheumatology, Department of Medicine, University of Alabama, 510 20th Street, FOT 805B South, Birmingham, AL 35294, USA; Jasvinder.md{at}gmail.com

Abstract

Objective To study 90-day complications following total hip arthroplasty (THA) or total knee arthroplasty (TKA).

Method In a population-based cohort of all Olmsted County residents who underwent a THA or TKA (1994-2008), we assessed 90-day occurrence and predictors of cardiac complications (myocardial infarction, cardiac arrhythmia or congestive heart failure), thromboembolic complications (deep venous thrombosis or pulmonary embolism) and mortality.

Results 90-day complication rates after THA and TKA were: cardiac, 6.9% and 6.7%; thromboembolic, 4.0% and 4.9%; and mortality, 0.7% and 0.4%, respectively. In multivariable-adjusted logistic regression analyses, ASA class III–IV (OR 6.1, 95% CI:1.6-22.8) and higher Deyo–Charlson comorbidity score (OR 1.2, 95% CI:1.0-1.4) were significantly associated with odds of 90-day cardiac event post-THA in patients with no known previous cardiac event. In those with known previous cardiac disease, ASA class III–IV (OR 4.4, 95% CI:2.0-9.9), male gender (OR 0.5, 95% CI:0.3-0.9) and history of thromboembolic disease (OR 3.2; 95% CI:1.4-7.0) were significantly associated with odds of cardiac complication 90 days post-THA. No significant predictors of thromboembolism were found in THA patients. In TKA patients with no previous cardiac history, age >65 years (OR 4.1, 95% CI:1.2-14.0) and in TKA patients with known cardiac disease, ASA class III–IV (OR 3.2, 95% CI:1.8-5.7) was significantly associated with odds of 90-day cardiac events. In TKA patients with no previous thromboembolic disease, male gender (OR 0.5, 95% CI:0.2-0.9) and higher Charlson index (OR 1.2, 95% CI:1.1-1.3) and in patients with known thromboembolic disease, higher Charlson index score (OR 1.2, 95% CI:1.1-1.4) was associated with odds of 90-day thromboembolic events.

Conclusion Older age, higher comorbidity, higher ASA class and previous history of cardiac/thromboembolic disease were associated with an increased risk.

https://doi.org/10.1136/ard.2010.148726

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Total knee arthroplasty (TKA) and total hip arthroplasty (THA) lead to significant improvements in pain, function and health-related quality of life.1 2 Cardiac and thromboembolic complications are not uncommon in the peri and postoperative period. As a significant number of elderly patients undergo these common procedures, it is important to study these complications and their predictors. Most estimates of cardiac and thromboembolic complications following THA or TKA have been obtained from single-centre studies or national/regional datasets, including Medicare,3 4 the national inpatient sample (NIS)5 6 and state-based discharge registries.7,–,9 None of these estimates are truly population based. In an analysis of Medicare data, 90-day cardiac and thromboembolic complications and mortality after primary TKA were 0.8%, 0.8% and 0.6%, respectively.4 Following THA, 30-day mortality was 2% overall, higher after hip fracture (6%) than other diagnoses (0.9–1.6%).3

    Very few well-designed studies, with adequate sample size and adjustment for important confounders using multivariable analyses, have examined predictors of cardiac and thromboembolic complications and mortality. Using the NIS sample from 1990 to 2004, the authors found that older age, male gender, larger hospital size, comorbidities, postoperative complications and payer type were associated with higher inpatient mortality in patients who underwent THA or TKA.5 6 In an analysis of Medicare data from 1983 to 1985, older age was a significant predictor of 30-day mortality after THA in multivariable-adjusted analyses.3 Basilico et al10 performed a case–control study of 209 cases with postoperative cardiac complications (including pulmonary embolism (PE)) compared with 209 controls without any complication following THA or TKA. A previous history of heart disease (arrhythmia, coronary artery disease, myocardial infarction (MI), congestive heart failure (CHF) or valvular heart disease), older age, revision and bilateral surgery were associated with significantly higher odds of 90-day cardiac complications postarthroplasty.10 Whereas these studies have provided insight into cardiopulmonary complications following THA and TKA, they have important limitations. The studies are limited in that complications were studied only during inpatient stay (NIS),5 6 in certain age groups (≥65 years for Medicare),3 or they combined cardiac and thromboembolic complications.10 With the exception of one study,10 data were claims based.

    Our aim was to use a population-based approach using the data from the Rochester epidemiology project (REP) to study these complications after THA and TKA. The REP is a rich resource to conduct population-based studies.11,–,13 The objectives of this study were to: (1) provide population-based estimates of cardiac and thromboembolic events and mortality up to 90 days after THA and TKA; (2) assess the predictors of these outcomes at 90 days; and (3) examine the time-trends in these outcomes.

    Methods

    Data sources and study population

    We used the data from the Mayo Clinic (Rochester) electronic clinical and administrative databases for this study. We merged the data from the REP13 and the prospective Mayo Clinic total joint registry.14 The REP is a population-based records linkage system that captures health care and outcomes of residents of Olmsted County, which is situated in southeastern Minnesota and is composed of approximately 124 000 people (2000 US census). The REP was developed in the 1960s and has been continuously supported by the NIH since then13; it has been used to study population-based epidemiology of rheumatic diseases15,–,21 and several other conditions22,–,26 in the USA. Previous studies have shown that in any 3-year period, over 90% of Olmsted County residents are examined at the Mayo Clinic healthcare system in Olmsted County.13 The closest competing medical centres are in Minneapolis, Minnesota (139 km to the north) and LaCrosse, Wisconsin (114 km to the east). The Mayo Clinic provides primary and secondary care to local residents. The total joint registry has been prospectively capturing every hip and knee arthroplasty procedure since the procedures were introduced in 1969 and 1971, respectively. Patients are routinely followed up 1, 2 and 5 years after hip and knee arthroplasty. For patients who miss their appointments, mailed questionnaires regarding outcomes and complications are requested along with radiographs. Patients not returning for follow-up and not responding to mailed letters/questionnaires are contacted by telephone call from trained, dedicated registry staff, who administer these questionnaires over the phone.

    The time period of interest for this study was from 1 January 1994 to 31 May 2008. This longer observation period was chosen to provide us with enough outcome events, have all predictors of interest in our databases and to avoid significant changes in definitions of cardiac and thromboembolic events that we may have encountered. From the list of all elective THA/TKA conducted at the Mayo Clinic, we identified patients residing in Olmsted County, using the zip code +4 information, to obtain a population-based sample. As all Olmsted County residents use the Mayo Clinic system for their health care, this allowed us to capture all complications following the index arthroplasty. There were 1744 THA patients and 1604 TKA patients. After excluding those with fracture as the underlying diagnosis to identify only those with elective surgeries, the final sample consisted of 1195 THA and 1604 TKA patients.

    Study outcomes

    The three study outcomes of interest were cardiac events, thromboembolic events and all-cause mortality within 90 days of the index surgery in Olmsted County residents during our study period of interest, 1994–2008. We extracted the diagnoses codes for cardiac and thromboembolic events from the Mayo Clinic medical indexing system using Mayo Clinic's hospital adaptation of the international code for diseases codes27 (available from 1935 to the present) and the International Classification of Diseases ninth version. These codes are valid and are collected for every in and outpatient at the Mayo Clinic.28 Cardiac events were defined as new occurrence of MI, CHF or cardiac arrhythmia, identified by new codes for these conditions (see supplementary table S1, available online only). We also extracted procedure codes for angioplasty with and without stent and coronary artery bypass grafting. As the addition of these procedure codes led to the identification of only 0.9% of additional events in hip and 0.6% of additional events in knee patients who underwent the procedures without documentation of diagnostic codes for cardiac or thromboembolic events, we restricted our analyses to only those with the hospital adaptation of the international code for diseases or the International Classification of Diseases ninth version codes (ICD-9). A thromboembolic event was defined as the occurrence of new codes for deep vein thrombosis (DVT) or PE (see supplementary table S1, available online only).

    Predictors of outcomes

    In univariate analyses, we considered the following variables: (1) age, categorised as 65 years or less and over 65 years; (2) gender; (3) previous underlying cardiac events (MI, CHF, or arrhythmia), yes/no; (4) previous underlying thromboembolic event (DVT or PE), yes/no; (5) body mass index (BMI) available from 1988, for each 5 unit increase (values over 80 and under 10 were set to missing); (6) comorbidity assessed by the Deyo–Charlson index, a validated comorbidity assessment consisting of a weighted sum of 17 comorbidities (including cardiac, pulmonary, renal, hepatic disease, diabetes, cancer, HIV, etc)29 30 with higher scores indicating more comorbidity, per 1 point increase; available from 1994; (7) American Society of Anesthesiologists (ASA) score, a validated measure of perioperative mortality and immediate postoperative morbidity, categorised as classes I–II versus III–IV31 32 (class I, normal healthy patient; class II, patient with mild systemic disease (with no functional limitation); class III, patient with severe systemic disease (with some functional limitation); class IV, patient with severe systemic disease that is constant threat to life or moribund patient). As there were only seven bilateral THA, all 2006 or later, this variable was not analysed for THA. As a result of a dramatic effect of pre-existing cardiac and thromboembolic disease on the occurrence of these events after THA and TKA, all analyses were performed separately for those with and without pre-existing cardiac and pre-existing thromboembolic disease. In particular, the predictors considered for respective multivariable analyses were as follows: (a) predictors of 90-day cardiac event: age, gender, ASA, BMI, Deyo–Charlson index and pre-existing thromboembolic disease; (b) predictors of 90-day thromboembolic event: age, gender, ASA, BMI, Deyo–Charlson index and pre-existing cardiac disease. Only predictors statistically significant in univariate analyses (p<0.05) were entered into multivariable regression models. As the 90-day mortality was low (<10 each for THA and TKA), we could not perform any analyses for predictors of mortality.

    Statistical analyses

    Descriptive statistics are reported as numbers (percentages) or mean (SD) as appropriate. Olmsted County residents were assessed for the occurrence of a cardiac event, thromboembolic event and all-cause mortality within 90 days of their initial joint arthroplasty. A 90-day cardiac event included a diagnosis of arrhythmia, MI, or CHF within 90 days of the THA/TKA. A 90-day thromboembolic event included a diagnosis of either PE or DVT within 90 days of THA/TKA.

    Separate logistic regression models were used for univariate and multivariable-adjusted analyses of 90-day cardiac events and 90-day thromboembolic events. A backward selection method was used to identify the significant variable in the multivariable models. Variables with significant associations with the outcome (p<0.05) in univariate analyses were entered into the multivariable models. Variables assessed for association with each event included: age (categorised as ≤65 vs >65 years), male gender, ASA score (categorised as I/II vs III/IV), BMI, Deyo–Charlson index, thromboembolic event before THA/TKA (assessment of 90-day cardiac event), and cardiac event before THA/TKA (assessment of 90-day thromboembolic event). OR and 95% CI, based on the logistic regression model estimates, are reported. Both univariate and multivariable-adjusted estimates are presented in tables, but only multivariable-adjusted estimates are described in the Results and Discussion, to avoid complexity.

    Time trends in cardiac and thromboembolic events were examined for the various study periods using a linear regression with the independent variable being the time of the orthopaedic surgery. Trends in BMI and comorbidity were analysed similarly. The α level was set at 0.05 for statistical significance. All analyses were performed using SAS version 9.2.

    Results

    Characteristics of the study populations

    We identified all Olmsted County residents who underwent elective THA (n=1195) or TKA (n=1604) at the Mayo Clinic during 1994–2008, after excluding patients with hip fracture. Cardiac and thromboembolic events and mortality were assessed for this population-based cohort.

    Olmsted County residents who underwent THA and TKA had mean ages of 67 and 68 years, 43% and 37% were male and the underlying diagnosis was osteoarthritis in 84% and 92%, respectively (table 1).

    View this table:
    Table 1

    Clinical characteristics of the study population

    Incidence of complications after THA and TKA

    Table 2 shows the population-based incidence of cardiac and thromboembolic events and mortality in THA and TKA in the Olmsted County cohort at 7, 30 and 90 days postoperatively. Overall, among patients from Olmsted County undergoing elective THA (n=1195), 6.9% had a cardiac event, 4.0% had a thromboembolic event and 0.7% died within 90 days post-THA. Similarly, among Olmsted County patients undergoing TKA (n=1604), 6.7% had a cardiac event, 4.9% had a thromboembolic event and 0.4% died within 90 days post-TKA. Table 3 shows the distribution of cardiac and thromboembolic events between patients who had previoius cardiac or previous thromboembolic events. The rates differed sharply between patients with and without previous cardiac and thromboembolic events. Therefore, all subsequent analyses were performed separately for patients with and without previous cardiac and thromboembolic events. As only eight THA and seven TKA patients died within 90 days of the respective procedure, no further analyses of predictors were performed for 90-day mortality.

    View this table:
    Table 2

    Frequency of cardiac events, thromboembolic events and mortality in Olmsted County residents

    View this table:
    Table 3

    90-Day cardiac and thromboembolic events based on previous similar events in Olmsted County residents

    Risk factors for cardiac and thromboembolic events 90 days after elective THA

    In multivariable-adjusted analyses of THA patients with no known previous cardiac disease, a higher (worse) ASA class of 3–4 and higher Deyo–Charlson index were associated with a significantly higher risk of a cardiac event 90 days post-THA (table 4). In multivariable-adjusted analyses of THA patients with known cardiac disease, a higher (worse) ASA class of 3–4, female gender and previous history of thromboembolic event were significantly associated with a higher risk of a cardiac event 90 days post-THA.

    View this table:
    Table 4

    Univariate and multivariable-adjusted 90-day event rates in Olmsted County residents after THA

    In THA patients with or without known previous thromboembolic disease, none of the risk factors were significantly associated with the 90-day risk of a thromboembolic event (table 4).

    Risk factors for cardiac and thromboembolic events 90 days after primary TKA

    In TKA patients with no known previous cardiac disease, age over 65 years was associated with a higher risk of a cardiac event 90 days post-TKA (table 5). In TKA patients with known cardiac disease, a higher ASA class of 3–4 was significantly associated with a higher risk of a cardiac event 90 days post-TKA.

    View this table:
    Table 5

    Univariate and multivariable-adjusted 90-day event rate in Olmsted County residents after TKA

    In TKA patients with no known previous thromboembolic disease, a higher Deyo–Charlson index score and female gender were associated with a higher 90-day risk of thromboembolic events (table 5). In TKA patients with known previous thromboembolic disease, a higher Deyo–Charlson index score was associated with a significantly higher risk of thromboembolic events within 90 days of index TKA.

    Time trends in postoperative cardiac events and thromboembolic events

    We noted significant time trends with increases in cardiac events in both in the THA and TKA cohorts (figure 1A,C). No significant time trends in patients with thromboembolic events were noted, in both the THA and TKA cohorts (figure 1B,D).

    Figure 1
    Figure 1

    Time trends in cardiac and thromboembolic events in Olmsted County residents. (A) 90-Day cardiac events in Olmsted county residents with total hip arthroplasty (THA). (B) 90-Day thromboembolic events in Olmsted county residents with THA. (C) 90-Day cardiac events in Olmsted county residents with total knee arthroplasty (TKA). (D) 90-Day thromboembolic events in Olmsted county residents with TKA.

    Interestingly, significant increaseswas noted in BMI (p<0.001), and the Deyo-Charlson index in the THA cohort during the study period (p=0.02). For example BMI increased from 26.5 kg/m2 in 1994–6 to 28.8 kg/m2 in 2006–8 and Deyo–Charlson index increased from 2.0 in 1994–6 to 2.6 in 2006–8 in the THA cohort. Similarly in the TKA cohort, BMI increased significantly over the study period (p<0.001), for example it increased from 30.3 kg/m2 in 1994–6 to 32.1 kg/m2 in 2006–8.

    Discussion

    In this population-based study, we report population-based estimates of cardiac and thromboembolic complications and mortality in a US cohort of elective THA and TKA for the first time. We found that female gender and higher ASA class of III–IV were risk factors for cardiac events after THA. In TKA patients, older age and ASA class of III–IV were risk factors for 90-day cardiac events, and female gender and higher comorbidity index were risk factors for 90-day thromboembolic events. With very few exceptions (cardiac events in those with a previous history of cardiac events), we did not note any significant time trends. An increase in disease diagnosis, closer surveillance for these complications and a change in disease definitions may contribute to time trends. Several findings deserve further discussion.

    This study provides population-based estimates for 90-day cardiac and thromboembolic events in a US arthroplasty cohort and adds to the literature. We know of no other population-based US study that has provided incidence rates for postarthroplasty complications. Overall, in the THA cohort, 6.9% had 90-day cardiac and 4.0% had 90-day thromboembolic events. In the TKA cohort, 6.7% had 90-day cardiac and 4.9% had 90-day thromboembolic events. It is not surprising that the rate of 1.4% for 90-day MI in our TKA cohort is very similar to the 0.8% reported for a primary TKA Medicare study population from the year 2000.4 One difference in these studies was the inclusion of all patients in our study versus only those 65 years and older in the previous study that used Medicare data.

    Although not directly comparable, the thromboembolic event rates from two previously published non-population-based US studies were 1.9% for THA and 3.0% for TKA patients during the index hospitalisation7 and 2.8% for elective THA and 2.1% for primary TKA within 90 days.8 Differences in study design (population vs non-population based), study time frames and mean age in the THA cohort may underlie these differences. Non-population-based studies lack the ability to provide true incidence rates, because the denominator is patients seen in a facility and not the population. Therefore, they can over or underestimate rates based on the patient case mix. Rates of symptomatic thromboembolic event 90 day postoperatively in a population-based study from Scotland were 2.2% for THA and 1.7% for TKA during 1992–2001.33 Differences in outcome event definition (any vs symptomatic thromboembolism), healthcare delivery systems (US vs Sweden), risk factor profile of patients, patient age (69 vs 71 years for THA) and study time period differences may explain the differences in the rates between our and the Swedish study.

    We identified risk factors for cardiac events following THA and TKA. An important observation was that the factors and the risk associated with them varied by the type of arthroplasty (THA vs TKA) and the history of previous cardiac or thromboembolic disease. In fact, the complication rates were so dramatically different between patients with and without pre-existing disease that all analyses were conducted separately. These observations imply that caution must be exercised before combining events in patients with and without a previous history of these events and in patients with THA versus TKA.

    One of the previous well-designed case–control studies combined THA and TKA cohorts and included PE in cardiovascular complications.10 In that study, a previous history of cardiac disease, older age, revision and bilateral surgery were associated with higher odds of 90-day cardiac complications after total joint arthroplasty.10 We confirm the previously reported old age and pre-existing cardiac disease associations with cardiac complications. In addition, not surprisingly, we identified ASA class as a risk factor for cardiac complications, in both the THA and TKA cohorts. ASA class has been identified as a risk factor for peri-operative complications in general.34 Our study extends this finding to postoperative cardiac complications in arthroplasty cohorts.

    We found that there were no additional demonstrable risk factors for 90-day thromboembolic events in THA patients except a previous history of thromboembolism (unadjusted rates 5.9% vs 20.3%). Our study confirms a similar finding of association of previous thromboembolic disease with 30-day postoperative symptomatic thromboembolic events in patients undergoing major lower limb surgery, including THA, TKA and hip fracture surgery,35 and extends this to 90-day follow-up. The surgeons may need to be extra vigilant for thromboembolic events in THA patients with previous thromboembolism with regard to the choice and intensity of venous thromboembolism prophylaxis. There is an intense debate regarding the choice of thromboprophylaxis in patients undergoing knee or hip arthroplasty.36 37 The American Association of Chest Physicians recommends the use of heparin, coumadin or fondaparinux and recommends not using aspirin or mechanical devices alone.38 On the other hand, the American Academy of Orthopaedic Surgeons recommends aspirin as an option in those at lower risk of thromboembolism and higher risk of bleeding.39

    Women undergoing TKA had a significantly higher risk of thromboembolism in our study, confirming a similar previous observation.7 In addition, we found that higher comorbidity was a significant risk factor for thromboembolic events in both patients with and without previous known thromboembolism. A similar observation has been made in THA, but not TKA, cohorts.7 Our study is limited in that we did not study post-thrombotic syndrome, an important complication after TKA/THA. DVT occurred more frequently than PE in both THA and TKA cohorts, which is similar to previous studies.40,–,44 This is an expected finding due to immobilisation of the lower extremity after THA/TKA.

    In conclusion, we performed a population-based study to estimate the incidence rates of cardiac and thromboembolic events and mortality up to 90 days after elective THA and TKA. Several findings of risk factors from other studies were confirmed and new predictors and risk factors were identified. Findings from this study can guide providers to discuss the risk of these complications depending on whether patients have a previous history of cardiac and thromboembolic disease or not. Future studies should explore whether the use of certain interventions, including the type of thromboprophylaxis and comorbidity management, can reduce these perioperative complications.

    References

    1. ↵
      1. Kane RL,
      2. Saleh KJ,
      3. Wilt TJ,
      4. et al
      . The functional outcomes of total knee arthroplasty. J Bone Joint Surg Am 2005;87:1719–24.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Ethgen O,
      2. Bruyère O,
      3. Richy F,
      4. et al
      . Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg Am 2004;86-A:963–74.
      OpenUrlPubMedWeb of Science
    3. ↵
      1. Whittle J,
      2. Steinberg EP,
      3. Anderson GF,
      4. et al
      . Mortality after elective total hip arthroplasty in elderly Americans. Age, gender, and indication for surgery predict survival. Clin Orthop Relat Res 1993;295:119–26.
      OpenUrlPubMed
    4. ↵
      1. Katz JN,
      2. Barrett J,
      3. Mahomed NN,
      4. et al
      . Association between hospital and surgeon procedure volume and the outcomes of total knee replacement. J Bone Joint Surg Am 2004;86-A:1909–16.
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Memtsoudis SG,
      2. Ma Y,
      3. González Della Valle A,
      4. et al
      . Perioperative outcomes after unilateral and bilateral total knee arthroplasty. Anesthesiology 2009;111:1206–16.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Memtsoudis SG,
      2. Della Valle AG,
      3. Besculides MC,
      4. et al
      . Risk factors for perioperative mortality after lower extremity arthroplasty: a population-based study of 6,901,324 patient discharges. J Arthroplasty 2010;25:19–26.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Lyman S,
      2. Jones EC,
      3. Bach PB,
      4. et al
      . The association between hospital volume and total shoulder arthroplasty outcomes. Clin Orthop Relat Res 2005;432:132–7.
      OpenUrlPubMed
    8. ↵
      1. White RH,
      2. Romano PS,
      3. Zhou H,
      4. et al
      . Incidence and time course of thromboembolic outcomes following total hip or knee arthroplasty. Arch Intern Med 1998;158:1525–31.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Kreder HJ,
      2. Deyo RA,
      3. Koepsell T,
      4. et al
      . Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington. J Bone Joint Surg Am 1997;79:485–94.
      OpenUrlPubMed
    10. ↵
      1. Basilico FC,
      2. Sweeney G,
      3. Losina E,
      4. et al
      . Risk factors for cardiovascular complications following total joint replacement surgery. Arthritis Rheum 2008;58:1915–20.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. St Sauver JL,
      2. Grossardt BR,
      3. Yawn BP,
      4. et al
      . Use of a medical records linkage system to enumerate a dynamic population over time: the Rochester epidemiology project. Am J Epidemiol 2011;173:1059–68.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Kremers HM,
      2. Myasoedova E,
      3. Crowson CS,
      4. et al
      . The Rochester Epidemiology Project: exploiting the capabilities for population-based research in rheumatic diseases. Rheumatology (Oxford) 2011;50:6–15.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Melton LJ III.
      . History of the Rochester Epidemiology Project. Mayo Clin Proc 1996;71:266–74.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Berry DJ,
      2. Kessler M,
      3. Morrey BF
      . Maintaining a hip registry for 25 years. Mayo Clinic experience. Clin Orthop Relat Res 1997;378:61–8.
      OpenUrl
    15. ↵
      1. Schäfer VS,
      2. Kermani TA,
      3. Crowson CS,
      4. et al
      . Incidence of herpes zoster in patients with giant cell arteritis: a population-based cohort study. Rheumatology (Oxford) 2010;49:2104–8.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Kermani TA,
      2. Schäfer VS,
      3. Crowson CS,
      4. et al
      . Malignancy risk in patients with giant cell arteritis: a population-based cohort study. Arthritis Care Res (Hoboken) 2010;62:149–54.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Doran MF,
      2. Crowson CS,
      3. O'Fallon WM,
      4. et al
      . Trends in the incidence of polymyalgia rheumatica over a 30 year period in Olmsted County, Minnesota, USA. J Rheumatol 2002;29:1694–7.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Shbeeb M,
      2. Uramoto KM,
      3. Gibson LE,
      4. et al
      . The epidemiology of psoriatic arthritis in Olmsted County, Minnesota, USA, 1982–1991. J Rheumatol 2000;27:1247–50.
      OpenUrlPubMedWeb of Science
    19. ↵
      1. Peterson LS,
      2. Nelson AM,
      3. Su WP,
      4. et al
      . The epidemiology of morphea (localized scleroderma) in Olmsted County 1960–1993. J Rheumatol 1997;24:73–80.
      OpenUrlPubMedWeb of Science
    20. ↵
      1. Peterson LS,
      2. Mason T,
      3. Nelson AM,
      4. et al
      . Juvenile rheumatoid arthritis in Rochester, Minnesota 1960–1993. Is the epidemiology changing? Arthritis Rheum 1996;39:1385–90.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Machado EB,
      2. Gabriel SE,
      3. Beard CM,
      4. et al
      . A population-based case–control study of temporal arteritis: evidence for an association between temporal arteritis and degenerative vascular disease? Int J Epidemiol 1989;18:836–41.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Gabriel SE,
      2. O'Fallon WM,
      3. Beard CM,
      4. et al
      . Trends in the utilization of silicone breast implants, 1964–1991, and methodology for a population-based study of outcomes. J Clin Epidemiol 1995;48:527–37.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Codd MB,
      2. Sugrue DD,
      3. Gersh BJ,
      4. et al
      . Epidemiology of idiopathic dilated and hypertrophic cardiomyopathy. A population-based study in Olmsted County, Minnesota, 1975–1984. Circulation 1989;80:564–72.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Melton LJ,
      2. 3rd,
      3. Crowson CS,
      4. O'Fallon WM
      . Fracture incidence in Olmsted County, Minnesota: comparison of urban with rural rates and changes in urban rates over time. Osteoporos Int 1999;9:29–37.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Bambha K,
      2. Kim WR,
      3. Talwalkar J,
      4. et al
      . Incidence, clinical spectrum, and outcomes of primary sclerosing cholangitis in a United States community. Gastroenterology 2003;125:1364–9.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Kim WR,
      2. Benson JT,
      3. Therneau TM,
      4. et al
      . Changing epidemiology of hepatitis B in a U.S. community. Hepatology 2004;39:811–16.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      Commission on Professional and Hospital Activities. Hospital Adaptation of ICDA. 2nd edn. Vol. 1. Ann Arbor, MI: Commission on Professional and Hospital Activities, 1973.
    28. ↵
      1. St Sauver JL,
      2. Hagen PT,
      3. Cha SS,
      4. et al
      . Agreement between patient reports of cardiovascular disease and patient medical records. Mayo Clin Proc 2005;80:203–10.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Charlson ME,
      2. Pompei P,
      3. Ales KL,
      4. et al
      . A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Charlson ME,
      2. Sax FL,
      3. MacKenzie CR,
      4. et al
      . Morbidity during hospitalization: can we predict it? J Chronic Dis 1987;40:705–12.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Dripps RD,
      2. Lamont A,
      3. Eckenhoff JE
      . The role of anesthesia in surgical mortality. JAMA 1961;178:261–6.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Weaver F,
      2. Hynes D,
      3. Hopkinson W,
      4. et al
      . Preoperative risks and outcomes of hip and knee arthroplasty in the Veterans Health Administration. J Arthroplasty 2003;18:693–708.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Howie C,
      2. Hughes H,
      3. Watts AC
      . Venous thromboembolism associated with hip and knee replacement over a ten-year period: a population-based study. J Bone Joint Surg Br 2005;87:1675–80.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Perka C,
      2. Arnold U,
      3. Buttgereit F
      . Influencing factors on perioperative morbidity in knee arthroplasty. Clin Orthop Relat Res 2000:183–91.
    35. ↵
      1. Leizorovicz A,
      2. Turpie AG,
      3. Cohen AT,
      4. et al
      . Epidemiology of venous thromboembolism in Asian patients undergoing major orthopedic surgery without thromboprophylaxis. The SMART study. J Thromb Haemost 2005;3:28–34.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Lieberman JR,
      2. Barnes CL,
      3. Lachiewicz PF,
      4. et al
      . Venous thromboembolism debate in joint arthroplasty. J Bone Joint Surg Am 2009;91(Suppl 5):29–32.
      OpenUrlCrossRefPubMed
    37. ↵
      1. Eikelboom JW,
      2. Karthikeyan G,
      3. Fagel N,
      4. et al
      . American Association of Orthopedic Surgeons and American College of Chest Physicians guidelines for venous thromboembolism prevention in hip and knee arthroplasty differ: what are the implications for clinicians and patients? Chest 2009;135:513–20.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Geerts WH,
      2. Bergqvist D,
      3. Pineo GF,
      4. et al
      . Prevention of venous thromboembolism: American College of Chest Physicians evidence-based clinical practice guidelines (8th Edn). Chest 2008;133(6 Suppl):381S–453S.
      OpenUrlCrossRefPubMedWeb of Science
    39. ↵
      American Academy of Orthopaedic Surgeons. Clinical guideline on prevention of symptomatic pulmonary embolism in patients undergoing total hip or knee arthroplasty. Rosemont, IL: American Academy of Orthopaedic Surgeons (AAOS), 2007.
    40. ↵
      1. Husted H,
      2. Otte KS,
      3. Kristensen BB,
      4. et al
      . Low risk of thromboembolic complications after fast-track hip and knee arthroplasty. Acta Orthop 2010;81:599–605.
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Lotke PA,
      2. Steinberg ME,
      3. Ecker ML
      . Significance of deep venous thrombosis in the lower extremity after total joint arthroplasty. Clin Orthop Relat Res 1994;299:25–30.
      OpenUrlPubMed
    42. ↵
      1. Kume H,
      2. Inoue Y,
      3. Mitsuoka A,
      4. et al
      . Doppler ultrasonography-aided early diagnosis of venous thromboembolism after total knee arthroplasty. Eur J Vasc Endovasc Surg 2010;40:664–8.
      OpenUrlCrossRefPubMed
    43. ↵
      1. Dushey CH,
      2. Bornstein LJ,
      3. Alexiades MM,
      4. et al
      . Short-term coagulation complications following total knee arthroplasty. A comparison of patient-reported and surgeon-verified complication rates. J Arthroplasty 2011;(in Press).
    44. ↵
      1. Westrich GH,
      2. Farrell C,
      3. Bono JV,
      4. et al
      . The incidence of venous thromboembolism after total hip arthroplasty: a specific hypotensive epidural anesthesia protocol. J Arthroplasty 1999;14:456–63.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Web Only Data

      Files in this Data Supplement:

    • Web Only Data

      Files in this Data Supplement:

    Footnotes

    • The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.

    • Funding This study received funding from the NIH CTSA Award 1 KL2 RR024151-01 (Mayo Clinic Center for Clinical and Translational Research) and the Department of Orthopaedic Surgery, Mayo Clinic School of Medicine, Rochester, Minnesota, USA.

    • Competing interests One of the authors (DGL) has received royalties/speaker fees from Zimmer, has been a paid consultant to Zimmer and has received institutional research funds from DePuy, Stryker and Zimmer. JAS has received speaker honoraria from Abbott; research and travel grants from Allergan, Takeda, Savient, Wyeth and Amgen, and consultant fees from Savient, URL pharmaceuticals and Novartis.

    • Ethics approval This study was conducted with the approval of the Mayo Clinic. Each author certifies that his or her institution has approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.

    • Provenance and peer review Not commissioned; externally peer reviewed.