Advertisement

Economic Burden of Surgical Management of Prosthetic Joint Infections Following Hip and Knee Replacements in Alberta, Canada: An analysis and comparison of two major urban centers

  • Kwadwo Mponponsuo
    Correspondence
    Corresponding Author: Room 419, North Tower, FMC, 1403 29th Street NW, Calgary, Alberta, T2N2T9. Phone: 639.999.3251.
    Affiliations
    Department of Medicine, University of Calgary, Calgary, Alberta, Canada
    Search for articles by this author
  • Jenine Leal
    Affiliations
    Department of Microbiology, Immunology and Infectious Diseases, University of Calgary, Canada

    Department of Community Health Sciences, University of Calgary, Calgary, AB, Canada

    O’Brien Institute for Public Health, University of Calgary, Calgary, AB, Canada

    Infection Prevention and Control, Alberta Health Services, AB, Canada
    Search for articles by this author
  • Shannon Puloski
    Affiliations
    Division of Orthopedic Surgery, Department of Surgery, University of Calgary, Calgary, AB, Canada
    Search for articles by this author
  • Derek Chew
    Affiliations
    Department of Medicine, University of Calgary, Calgary, Alberta, Canada

    Department of Community Health Sciences, University of Calgary, Calgary, AB, Canada

    O’Brien Institute for Public Health, University of Calgary, Calgary, AB, Canada

    Libin Cardiovascular Institute, Department of Cardiac Sciences, University of Calgary, Calgary, AB, Canada
    Search for articles by this author
  • Swati Chavda
    Affiliations
    Department of Medicine, University of Calgary, Calgary, Alberta, Canada
    Search for articles by this author
  • Arif Ismail
    Affiliations
    Department of Medicine, University of Alberta, Edmonton, Alberta, Canada
    Search for articles by this author
  • Flora Au
    Affiliations
    Department of Medicine, University of Calgary, Calgary, Alberta, Canada
    Search for articles by this author
  • Elissa Rennert-May
    Affiliations
    Department of Medicine, University of Calgary, Calgary, Alberta, Canada

    Department of Microbiology, Immunology and Infectious Diseases, University of Calgary, Canada

    Department of Community Health Sciences, University of Calgary, Calgary, AB, Canada

    O’Brien Institute for Public Health, University of Calgary, Calgary, AB, Canada

    Snyder Institute for Chronic Diseases, University of Calgary, Calgary, AB, Canada
    Search for articles by this author

      STRUCTURED SUMMARY

      Background

      Complex surgical site infections (SSIs) and revisions for these infectious complications following total knee and hip arthroplasties are associated with significant economic costs.

      Aim

      To evaluate the cost of one-stage and two-stage revisions; debridement, antibiotic, and implant retention (DAIR) and DAIR with liner exchange for complex hip or knee SSIs in Alberta, Canada.

      Methods

      We used the Alberta Health Services Infection Prevention and Control database to identify individuals >18 years old from the two major urban centers in Alberta, Calgary, and Edmonton zone, with complex hip or knee SSIs who underwent surgical intervention between April 1, 2012, and March 31, 2019. Micro-costing and gross-costing methods were used to estimate 12 and 24-month costs following the initial hospital admission for arthroplasty. Subgroup, inverse gaussian and gamma regression analysis were used to evaluate the associations of the revision procedure, age, sex, and comorbidities on cost.

      Findings

      A total of 382 patients with complex SSIs were identified with a mean age of 66.1 years. DAIR and DAIR with liner exchange resulted in the lowest 12- and 24-month costs at $53,197 (95% CI, $38,006 - $68,388) and $57,340 (95% CI, $48,576 - $66,105), respectively; two-stage revision was the costliest procedure. Most of the incurred costs (>98%) were accrued within the first 12 months following the initial procedure.

      Conclusions

      Medical costs are highest 12 months following initial arthroplasty and for two-stage revisions in hip and knee complex SSI.

      Key Words

      Introduction

      Total hip and knee arthroplasties are common surgical procedures that can ameliorate pain and improve mobility. Between 2019 and 2020, despite reductions in surgeries as a result of the COVID-19 global pandemic, over 137,000 of these surgeries were performed, accounting for over $1.4 billion in inpatient hospital costs in Canada [

      Canadian Institute for Health Information. Hip and Knee Replacements in Canada, 2017–2018: Canadian Joint Replacement Registry Annual Report. Ottawa, ON: CIHI; 2019.

      ]. Arthritis is the most common reason for these procedures with an estimated 6 million Canadians currently living with the condition. By 2040, that number is expected to increase to 9 million []. Thus, the number of total hip and knee arthroplasties conducted will concurrently increase. While these procedures improve quality of life, severe post-operative infectious complications, which occur in 1-2% of individuals, result in increased morbidity, mortality, and extensive healthcare costs [
      • Miletic K.G.
      • Taylor T.N.
      • Martin ET Vaidya R.
      • Kaye K.S.
      Readmissions after diagnosis of surgical site infection following knee and hip arthroplasty.
      ,
      • Nair R.
      • Schweizer M.L.
      • Singh N.
      Septic Arthritis and Prosthetic Joint Infections in Older Adults.
      ,
      • Triffault-Fillit C.
      • Ferry T.
      • Laurent F.
      • et al.
      Microbiologic epidemiology depending on time to occurrence of prosthetic joint infection: a prospective cohort study.
      ].
      Management of post-operative infections is dependent on the classification of infection; superficial infections are managed with antibiotics with or without local debridement, while more extensive surgical intervention combined with antimicrobial therapy is needed for those with deep incisional and organ-space disease (i.e. complex infections) [

      American Academy of Orthopaedic Surgeons EvidenceBased Clinical Practice Guideline for Diagnosis and Prevention of Periprosthetic Joint Infections. https://www.aaos.org/pjiguideline Published March 11, 2019

      ]. Debridement with antibiotics and implant retention (DAIR), DAIR with liner exchange, one-stage revision and two-stage revision are the four surgical strategies used for the management of complex surgical site infections (SSI). Clinical cure has been variably reported with rates of 86.9% to 91.8% described for one-stage revisions and 85% to 92.1% for two-stage revisions [
      • Lange J.
      • Troelsen A.
      • Thomsen R.W.
      • Søballe K.
      Chronic infections in hip arthroplasties: comparing risk of reinfection following one-stage and two-stage revision: a systematic review and meta-analysis.
      ,
      • Silvestre A.
      • Almeida F.
      • Renovell P.
      • Morante E.
      • López R.
      Revision of infected total knee arthroplasty: two-stage reimplantation using an antibiotic-impregnated static spacer.
      ,
      • Kunutsor S.K.
      • Whitehouse M.R.
      • Blom A.W.
      • Beswick A.D.
      Re-Infection Outcomes following One- and Two-Stage Surgical Revision of Infected Hip Prosthesis: A Systematic Review and Meta-Analysis.
      ]. Lower rates of success have generally been reported for DAIR and DAIR with liner exchange at 31% - 100% [
      • Tsang S.J.
      • Ting J.
      • Simpson A.
      • Gaston P.
      Outcomes following debridement, antibiotics and implant retention in the management of periprosthetic infections of the hip: a review of cohort studies.
      ,
      • Qasim S.N.
      • Swann A.
      • Ashford R.
      The DAIR (debridement, antibiotics and implant retention) procedure for infected total knee replacement - a literature review.
      ]. As demonstrated by our previous work evaluating the total medical costs of these procedures in Calgary, Alberta, Canada (one of the two largest health zones within Alberta Health Services), these procedures have inherent differences in cost due to differences in resource utilization, duration of hospitalization and peri-operative complications. Our initial study revealed two-stage revisions to be the costliest intervention at $100,992 (95% CI, $34,587–$167,396) with a majority of the costs accruing in the first 12 months following revision intervention [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. To gain a more comprehensive understanding of the full economic impact of these interventions in the entire province we first conducted an analysis of the Edmonton zone (the second of the two largest health zones in Alberta Health Services) and then combined the Edmonton and Calgary zones within Alberta Health Services (AHS) to provide an updated cohort and to obtain a closer approximation of the true costs of the revision procedures.

      Methods

      Overview

      We conducted a population-based cohort study of patients undergoing surgical intervention for complex hip or knee SSI in Edmonton, Alberta, Canada. We then combined the Edmonton cohort with data from Calgary, Alberta, Canada to produce the combined cohort. Edmonton has a population of approximately 1.49 million people and is serviced by the single-healthcare system of Alberta, Alberta Health Services (AHS). After Calgary, Edmonton is the second largest city in Alberta. The combined population of Edmonton and Calgary totals approximately 3.07 million people, accounting for 69.1% of the entire population of Alberta [

      Government of Alberta. (2021). Population Projects: Alberta and Census Divisions, 2021–2046. https://www.alberta.ca/population-statistics.aspx#data.

      ].

      Patient cohort

      We identified a cohort of adults 18 years or older who were admitted to a hospital in the Edmonton zone for complex SSI of a prosthetic hip or knee within 90-days of the primary arthroplasty between April 1, 2012, and March 31, 2019. We subsequently combined the Edmonton zone cohort with the Calgary zone data from our prior study to create a combined cohort, representing a majority of prosthetic joint infection revisions in Alberta for analysis. Of note, a prior study evaluating the incidence of complex SSI in Alberta yielded a rate of 1.04% (1.48% for hips and 0.76% for knees) [
      • Rennert-May E.D.
      • Conly J.
      • Smith S.
      • et al.
      The cost of managing complex surgical site infections following primary hip and knee arthroplasty: A population-based cohort study in Alberta, Canada.
      ]. The cohort was then separated based on the surgical intervention received for management of their complex SSI (i.e., DAIR, DAIR with liner exchange, one-stage revision, and two-stage revision). The type of surgical management was identified through chart review for all patients identified in the base cohort. The same process was previously utilized to identify the Calgary cohort [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ].

      Data Sources

      Our cohort was extracted from the AHS Infection Prevention and Control (IPC) surveillance group database which collects, verifies, and stores data on all complex SSIs that occur within ninety days following hip and knee replacements in Alberta.
      Following extraction of our cohort from the AHS IPC surveillance database, we utilized the Discharge Abstract Database (DAD) and National Ambulatory Care Reporting System (NACRS), respectively, from AHS analytics to identify hospital admissions and outpatient visits to AHS facilities occurring within 12 and 24 months following the date of the initial arthroplasty.
      Patients with infection identified in 2019 were only followed for 12 months for all outcomes due to availability of data. Elixhauser comorbidities were also defined from DAD for all patients.

      Costing data

      Micro-costing and gross-costing methods were utilized to capture the costs associated with the healthcare utilization of each patient in the cohort. Micro-costing, considering the gold standard of costing data, provides specific patient levels costs where each component of resource use is estimated and a unit cost is derived for each [
      • Clement Nee Shrive F.M.
      • Ghali W.A.
      • Donaldson C.
      • Manns B.J.
      The impact of using different costing methods on the results of an economic evaluation of cardiac care: microcosting vs gross-costing approaches.
      ]. AHS corporate finance provided micro-costing data for all inpatient costs (e.g., surgical procedures). As micro-costing was not available for outpatient visits and their associated costs, gross-costing methods were estimated from AHS analytics. Gross costing is considered a “top-down” approach where an average aggregated cost of healthcare usage is utilized. To do this, an average cost was calculated using the cost per standard hospital stay for Alberta and then multiplied by a person’s resource intensity weight, the standardized measure of the hospital resources utilized by an individual [

      Canadian Institute for Health Information. DAD resource intensity weights and expected length of stay (ELOS) for CMG+ 2020, https://secure.cihi.ca/estore/productSeries.htm?pc=PCC90; 2020 [accessed 20 January 2022].

      ]. Costs were collected from the Canadian public healthcare payer perspective.

      Outcomes

      The primary outcomes were 12- and 24-month cumulative direct healthcare cost from time of initial arthroplasty, including admission for complex SSI, subsequent hospitalizations and emergency room visits, and outpatient clinic visit costs. Physician billing costs were not included. As a secondary outcome we assessed total length of stay (LOS) in hospital at 12 and 24 months. All costs were inflated to 2020 Canadian dollars (CAD) using the consumer price index [

      Statistics Canada. Consumer Price Index: Annual Review, 2019. https://www150.statcan.gc.ca/n1/daily-quotidien/200122/dq200122c-eng.htm. Accessed November 2, 2020.

      ].

      Statistical Analyses

      We determined the mean cost and LOS for the initial admission and revision procedure in addition to all patient hospitalizations, emergency room visits, and outpatient visits in the subsequent 12 and 24 months. The cumulative mean 12- and 24-month cost was calculated for each of the four revision strategies for the two cohorts (i.e., DAIR, DAIR with liner exchange, one-stage, and two stage procedures) that included costs of the procedure itself. As an additional analysis, twelve-month means were also calculated by combining DAIR and DAIR with liner exchange into a single group, denoting procedures without hardware removal, along with a combined one-stage and two-stage revision group, denoting procedures with hardware removal. If an individual had a revision procedure for a complex SSI followed by a second revision for a failed revision (e.g., a DAIR followed by two-stage revision) within the total 24-month period analyzed, all costs accrued, including the cost of the second revision, were reflected in the initial revision cohorts. For those who had a complex SSI in 2019, only the first 12 months (until December 31, 2019) were included for analysis due to data availability. A negative binomial regression was utilized to calculate the 95% confidence intervals (CI) for all unadjusted costs due to overdispersion of data. Subgroup analyses assessed the 12- and 24-month costs by age (i.e., < 65 or > 65 years), and Elixhauser comorbidity index (i.e, 0, 1-2, > 3). Inverse gaussian distribution and gamma regression with log link analyses were performed to determine adjusted costs controlling for the type of revision procedure, age, sex, diabetes, hypertension with and without chronic complications, number of Elixhauser comorbidities (similar to the previous study) and disease caused by Staphylococcus aureus versus non-Staphylococcus aureus species [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. All statistical analyses were conducted using STATA/MP 16.0 (Texas, USA).
      The University of Calgary Health Research Ethics Board approved this study.

      Results

      Patient characteristics for the Edmonton cohort

      A total of 234 patients (51.3% male), with a mean age of 65.9 ± 11.1 [standard deviation (SD)], were identified. Elixhauser comorbidity scores of 0, 1 to 2 and >3 was observed in 41.5%, 32.9% and 25% of patients, respectively. Complex SSIs of the hip accounted for a majority (62.9%) of the initial infections. DAIR with liner exchange was the most common procedure performed (63.4%). Five patients did not receive any surgical intervention. The full demographic characteristics of the cohort can be found in Table 1.
      Table IBaseline characteristics of the Edmonton cohort
      PATIENT CHARACTERISTICSTotalDAIRDAIR with Liner ExchangeOne-stage exchangeTwo-stage exchangeNo surgical managementP value
      N of patient, N (%)23451 (21.8)146 (62.4)5 (2.1)27 (12.1)5 (2.1)
      Age (years), mean (+/-SD)65.9 +/- 11.167.6 +/- 10.065.9 +/- 11.359.6 +/- 17.063.4 +/- 9.169.2 +/- 17.30.17
      Male, N (%)120 (51.3)24 (47.1)76 (52.1)1 (20.0)17 (60.7)2 (40.0)0.39
      Death, N (%)22 (9.4)4 (7.8)13 (8.9)0 (0.0)4 (14.3)1 (20.0)0.68
      BMI, mean (+/-SD)34.5 +/- 7.833.5 +/- 6.434.9 +/- 7.931.3 +/- 8.235.2 +/- 8.435.5 +/- 12.80.17
      Elixhauser comorbidities
      0, N (%)97 (41.5)16 (31.4)66 (45.2)2 (40.0)11 (40.7)2 (40.0)0.56
      1 or 2, N (%)77 (32.9)21 (41.2)47 (32.2)2 (40.0)6 (22.2)1 (20.0)0.48
      >=3, N (%)60 (25.0)14 (26.4)33 (22.2)1 (20.0)10 (35.7)2 (40.0)0.52
      Select Comorbidities
      Diabetes without complications, N (%)21 (8.8)5 (9.8)11 (7.5)0 (0.0)5 (18.5)0 (0.0)0.35
      Hypertension without complications, N (%)84 (35.9)20 (39.2)52 (35.6)1 (20.0)9 (33.3)2 (40.0)0.92
      Revision Procedures
      Number of procedures, N (%)24053 (22.1)149 (62.1)5 (2.1)28 (11.7)5 (2.1)
      Time to first complex SSI, mean (+/-SD)24.4 +/- 15.522.4 +/- 14.022.3 +/- 12.917.0 +/- 10.939.5 +/- 20.137.6 +/- 29.0<0.001
      ASA score ≥ 3, N (%)119 (49.5)28 (52.8)72 (48.3)2(40.0)13 (46.4)4 (80.0)0.44
      DAIR: debridement, antibiotics, implant retention, BMI: body mass index; column percentages presented

      Patient characteristics for the combined cohort

      When the Edmonton zone and Calgary zone cohorts were combined, a total of 329 patients (53.4% male) with a mean age of 66.1 ± 10.9 (SD), were identified. Elixhauser comorbidities scores of 0, 1 or 2, and >3 was found in 42.5%, 32.8%, and 35.2% of patients, respectively. The most common procedure performed was DAIR with liner exchange occurring in 207 patients (62.9%); six patients did not undergo any surgical intervention (Table 2).
      Table IIBaseline characteristics of the combined cohort
      PATIENT CHARACTERISTICSTotalDAIRDAIR with Liner ExchangeOne-stage exchangeTwo-stage exchangeNo surgical managementP value
      Number of patients, N (%)32968 (20.7)207 (62.9)14 (4.3)34 (10.3)6 (1.8)
      Age (years),

      mean (+/-SD)
      66.1 +/- 10.966.8 +/- 9.766.0 +/- 11.269.5 +/- 13.263.6 +/- 8.966.3 +/- 17.10.07
      Male, N (%)204 (53.4)41 (52.6)127 (52.1)10 (55.6)23 (63.9)3 (50.0)0.76
      Death, N (%)35 (9.2)6 (7.7)20 (8.2)1 (5.6)7 (19.4)1(16.7)0.20
      BMI,

      mean (+/-SD)
      34.1 (7.9)33.0 (7.4)34.3 (7.7)34.3 (8.7)35.0 (8.9)35.5 (12.8)0.68
      Elixhauser comorbidities
      0, N (%)138 (42.0)26 (38.2)89 (43.0)9 (64.3)11 (32.4)3 (50.0)0.309
      1 or 2, N (%)108 (32.8)23 (33.8)72 (34.8)3 (21.4)9 (26.5)1 (16.7)0.636
      >=3, N (%)83 (25.2)19 (27.9)46 (22.2)2 (14.3)14 (41.2)2 (33.3)0.139
      Select Commodities
      Diabetes without complications, N (%)33 (10.0)9 (13.2)18 (8.7)0 (0.0)6 (16.7)0 (0.0)0.231
      Hypertension without complications, N (%)127 (38.6)28 (41.2)79 (38.2)2 (14.3)16 (47.1)2 (33.3)0.309
      Revision Procedures
      N of procedure, N (%)38278 (20.4)244 (63.9)18 (4.7)36 (9.4)6 (1.6)
      ASA score ≥ 3, N (%)162 (42.4)34 (43.6)98 (40.2)10 (55.6)15 (41.7)5 (83.3)0.217
      Time to first complex SSI, mean (+/-SD)25.0 +/- 16.322.7 +/- 14.623.7 +/- 15.018.1 +/- 9.241.1 +/- 19.835.2 +/- 26.60.002

      Base costs for the Edmonton cohort

      The overall mean 12- and 24-month costs were $66,829 (95% CI, $57,393 – $76,264) and $67,389 (95% CI, $45,876 – $76,901), respectively. One-stage revision accounted for the highest 12-month costs at $148,101 (95% CI, $21,367 – $274,835) with no additional cost incurred at 24-months. The mean 12-month cost for DAIR and DAIR with liner exchange combined was $59,365 (95% CI, $50,273 - $68,856). Higher mean 12-month cost was observed when one-stage and two-stage revisions were combined [$117,350 (95% CI, $82,422 – 152,277)] (data not shown). Inpatient costs for one-stage revision accounted for 70.3% of the overall 12-month cost. The operative procedure accounted for 3.7% of the overall 12-month cost. At 12-months following surgical intervention, DAIR and DAIR with liner exchange were the least costly interventions at $52,124 (95% CI, $37,040 – $67,208) and $62,212 ($50,749 – $73,675), respectively. Inpatient costs were responsible for 87.2% and 91.7% of the 12-month costs for DAIR and DAIR with liner exchange, respectively. The operative revision procedures accounted for less than 10% of the total 12-month costs for both DAIR and DAIR with liner exchange. DAIR and DAIR with liner exchanged remained the least costly interventions at 24-months, while one-stage revisions remained the highest costing procedure (Table 3).
      Table IIITotal mean costs, length of stay and number of hospitalizations at 12 and 24 months for the Edmonton cohort
      TotalDAIRDAIR with Liner ExchangeOne-stage RevisionTwo-stage RevisionNo Surgical Management
      N240531495285
      12-month overall cost ($)66,829

      (57,393 – 76,264)
      52,124

      (37,040 – 67,208)
      62,212

      (50,749 – 73,675)
      148,101

      (21,367 – 274,835)
      111,858

      (77,515 – 146,201)
      26,846

      (7,361 – 46,330)
      12-month hospital admissions (N)11.4

      (9.7 – 13.1)
      9.0

      (6.6 – 11.5)
      10.6

      (8.5 – 12.7)
      17.2

      (0.03 – 34.4)
      19.5

      (12.6 – 26.4)
      9.8

      (1.8 – 17.8)
      12-month LOS (days)40.5

      (32.6 – 48.4)
      30.0

      (18.8 – 41.2)
      37.7

      (27.3 – 48.1)
      97.0

      (2.2 – 196.2)
      69.5

      (42.2 – 96.7)
      18.8

      (4.9 – 32.7)
      24-month overall cost ($)67,389

      (57,876 – 76,901)
      52,713

      (37,202 – 68,225)
      62,553

      (51,043 – 74,064)
      148,101

      (21,367 – 274,835)
      113,701

      (79,109 – 148,293)
      26,981

      (7,633 – 46,328)
      24-month hospital admissions (N)11.5

      (9.8 – 13.2)
      9.2

      (6.6 – 11.7)
      10.7

      (8.6 – 12.7)
      17.2

      (0.03 – 34.4)
      19.9

      (12.8 – 27.0)
      9.8

      (1.8 – 17.8)
      24-month LOS (days)40.7

      (32.78 – 48.7)
      30.2

      (18.7 – 41.7)
      37.8

      (27.3 – 48.2)
      97.0

      (2.2 – 196.2)
      70.2

      (42.9 – 97.5)
      18.8

      (4.9 – 32.7)
      Inpatient Costs
      N233511445285
      12-month hospital cost ($)61,248

      (51,752 – 70,744)
      45,438

      (31,744 – 59,130)
      57,042

      (45,356 – 68,730)
      104,170

      (69,725 – 138,614)
      21,793

      (1,897 – 41,689)
      142,728

      (11,250 – 274,206)
      N412100
      24-month additional hospital cost ($)21,205

      (2,853 – 39,557)
      14,48115,939

      (8,756 – 23,122)
      38,460--
      Total 24-month hospital cost ($)61,613

      (52,054 – 71,170)
      45,722

      (31,700 – 59,750)
      57,264

      (45,540 – 68,988)
      105,543

      (70,824 – 140,262)
      21,793

      (1,897 – 41,690)
      142,728

      (11,250 – 274,206)
      Outpatient Costs
      N240531495285
      12-month outpatient cost ($)7,398

      (6,496 – 8299)
      8,400

      (5,541 – 11,260)
      7,131

      (6,151 – 8,111)
      5,374

      (2,297 – 8,450)
      7,689

      (5,580 – 9,798)
      5,053

      (1,391 – 8,715)
      N6818340133
      24-month additional outpatient cost ($)730

      (477 – 983)
      932

      (336 – 1,528)
      560

      (334 – 786)
      01,011

      (30 – 1,991)
      225

      (68 – 382)
      Total 24-month cost ($)7,605

      (6,670 – 8,541)
      8,717

      (5,750 – 11,685)
      7,260

      (6,251 – 8,268)
      5,374

      (2,297 – 8,450)
      8,158

      (5,894 – 10,423)
      5,187

      (1,506 – 8,869)

      Base costs for the combined cohort

      The overall mean 12- and 24-month cost was $62,224 (95% CI, $54,815 – $68,633) and $62,854 (95% CI, $55,407 – $70,300), respectively. Two-stage revisions resulted in the highest cost at 12 and 24 months at $114,030 (95% CI, $83,959 – $144,110) and $116,170 (95% CI, $91,225 – $141,115), respectively. The mean 12-month cost for DAIR and DAIR with liner exchange combined was $56,337 (95% CI, $48,781 - $63,892). Higher mean 12-month cost was observed when one-stage and two-stage revisions were combined [$101,671 (95% CI, $76,908 – 126,433)] (data not shown). Inpatient costs accounted for 96.0% of 12-month overall cost with the revision procedure accounting for 9.0% of the total 12-month costs. Similar to the Edmonton cohort base costs, DAIR and DAIR with liner exchange yielded the lowest 12- and 24-month costs (Table 4). Inpatient costs accounted for 87.9% and 89.7% of the total 12-month costs for DAIR and DAIR with liner exchange, respectively. Operative revisions costs accounted for less than 10% of the total 12-month costs of both DAIR and DAIR with liner exchange.
      Table IVTotal mean costs, length of stay and number of hospitalizations at 12 and 24 months for the combined cohort
      TotalDAIRDAIR with Liner ExchangeOne-stage RevisionTwo-stage RevisionNo Surgical Management
      N3827824418366
      12-month overall cost ($)62,224

      (54,815 – 69,633)
      53,197

      (38,006 – 68,388)
      57,340

      (48,576 – 66,105)
      76,954

      (31,299 – 122,609)
      114,030

      (83,959 – 144,100)
      23,146

      (5,612 – 40,479)
      12-month hospital admissions (N)10.5

      (9.2 – 11.8)
      8.5

      (6.6 – 10.4)
      9.6

      (8.2 – 11.1)
      10.5

      (4.5 – 16.5)
      21.1

      (1.5 – 28.7)
      8.2

      (0.8 – 15.5)
      12-month LOS (days)36.5

      (30.7 – 42.3)
      28.4

      (19.7 – 37.1)
      33.7

      (26.3 – 41.0)
      44.1

      (10.8 – 77.5)
      72.7

      (47.4 – 98.0)
      15.7

      (2.4 – 28.9)
      24-month overall cost ($)62,854

      (55,407 – 70,300)
      53,808

      (40,879 – 66,737)
      57,789

      (50,429 – 65,150)
      77,250

      (39,633 – 114,867)
      116,170

      (91,225 – 141,115)
      23,324

      (9,695 – 36,953)
      24-month hospital admissions (N)10.7

      (9.3 – 12.0)
      8.7

      (6.7 – 10.7)
      9.7

      (8.2 – 11.2)
      10.8

      (4.8 – 16.8)
      21.6

      (13.9 – 29.2)
      8.2

      (0.8 – 15.5)
      24-month LOS (days)36.8

      (30.9 – 42.7)
      29.2

      (19.8 – 38.6)
      33.8

      (26.5 – 41.2)
      44.1

      (10.8 – 77.5)
      73.6

      (48.4 – 98.8)
      15.7

      (2.4 – 28.9)
      Inpatient Costs
      N3737623918355
      12-month hospital cost ($)56,438

      (48,955 – 63,922)
      46,749

      (31,968 – 61,530)
      51,439

      (42,540 – 60,339)
      70,185

      (24,098 – 116,271)
      109,496

      (76,046 – 139,945)
      21,793

      (1,897 – 41,690)
      N1025120
      24-month additional hospital cost ($)15,970

      (8,975 – 22,966)
      14,420

      (13,658 – 15,183)
      13,561

      (10,356 – 16,766)
      73331,162

      (24,100 – 38,224)
      -
      Total 24-month hospital cost ($)56,715

      (49,215 – 64,215)
      47,128

      (32,125 – 62,132)
      51,723

      (42,811 – 60,634)
      70,226

      (24,146 – 116,305)
      108,185

      (78,059 – 138,312)
      21,793

      (1,897 – 41,690)
      Outpatient Costs
      N3827824418366
      12-month outpatient cost ($)7,133

      (6,505 – 7,763)
      7,647

      (5,512 – 9,782)
      6,984

      (6,322 – 7,645)
      6,769

      (4,855 – 8,684)
      7,576

      (5,838 – 9,313)
      4,985

      (2,211 – 7,758)
      N11126576184
      24-month additional outpatient cost ($)728.7

      (549.0 – 908.4)
      723.4

      (302.7 – 1,144)
      731.4

      (519.6 – 943.2)
      764.5

      (180.2 – 1,348.7)
      818.3

      (117.6 – 1,519.1)
      267.3

      (109.2 – 425.4)
      Total 24-month cost ($)7,346.1

      (6,700 – 7,996)
      7,887

      (5,683.7 – 10,092)
      7,155.3

      (6,473.9 – 7,837)
      7,024

      (5,093.0 – 8,955)
      7,985

      (6,121 – 9,848)
      5,162

      (2,379 – 7,947)

      Length of stay for the Edmonton cohort

      The average mean LOS was 40.5 days (95% CI, 32.6 days – 48.4 days) at 12-months with a slight increase at 24-months to 40.7 days. Twelve-months following the procedures, one-stage revisions had the highest LOS at 97.0 days (95% CI, 2.2 days – 196.2 days) while DAIR had the lowest at 30.0 days (95% CI, 18.8 days – 41.2 days). There was no incremental increase in LOS for one-stage revisions at 24 months and DAIR had minimal increase in LOS from 30.0 days to 30.2 days (Table 4).

      Length of stay for the combined cohort

      At 72.7 days (95% CI, 47.4 days – 98.0 days), two-stage revisions resulted in the longest LOS 12 months following the procedure, while DAIR resulted in the shortest [28.4 days (95% CI, 19.7 days – 37.1 days)]. The average LOS for the combined cohort was 36.5 days (95% CI, 30.7 days – 42.3 days). There were minimal increases in the LOS for all the procedures performed at 24-months (Table 4).

      Subgroup analyses and adjusted costs for the Edmonton cohort

      Subgroup analyses revealed those >65 year of age had higher overall costs at 12 and 24-months for DAIR with liner exchange and one-stage revisions (Table A.1 and Table A.2). There were greater costs associated with patients with greater Elixhauser comorbidities. For example, compared to patients with an Elixhauser score of 0, costs were $24,466 (p-value = 0.041) higher for patients with Elixhauser scores of 1 or 2 by Gamma regression. The association between increased costs and comorbidities was similar using inverse gaussian regression [$28,208 (p-value = 0.039) higher for those with Elixhauser comorbidity scores of 1 or 2]. Last, those >65 (compared to those <65 years) had significantly higher costs by inverse gaussian and gamma regression (Table 5).
      Table VGamma and Inverse Gaussian regression adjusted costs for the Edmonton cohort
      Gamma regressionInverse gaussian regression
      Coefficient ($)P-valueCoefficient ($)P-value
      DAIR with liner exchange (vs. DAIR)-11,5380.287DAIR with liner exchange (vs. DAIR)-12,7230.232
      One-stage (vs. DAIR)49,8810.106One-stage (vs. DAIR)41,1720.313
      Two-stage (vs. DAIR)41,9400.003Two-stage (vs. DAIR)49,0590.014
      Age ≥ 65 (vs <65)18,9680.036Age ≥ 65 (vs <65)21,6600.024
      Male (vs female)9,1340.309Male (vs female)13,3620.158
      ComorbiditiesComorbidities
      Diabetes without chronic complications (vs with complications)-9,9270.549Diabetes without chronic complications (vs with complications)-18,5630.272
      Hypertension without complications (vs with complications)-13,6330.336Hypertension without complication (vs with complications)-15,3070.336
      Elixhauser Index 1 or 2 (vs. 0)24,4660.041Elixhauser Index 1 or 2 (vs. 0)28,2080.039
      Elixhauser Index ≥ 3 (vs. 0)30,1740.088Elixhauser Index ≥ 3 (vs. 0)29,7810.121
      Staphylococcus aureus (vs. non-Staphylococcus aureus)7,5070.427Staphylococcus aureus (vs. non-Staphylococcus aureus)9,4850.352

      Subgroup analyses and adjusted costs for the combined cohort

      Differing from the Edmonton cohort, subgroup analyses of the combined cohort revealed those less than 65 years of age had the highest overall 12- and 24-month cost for DAIR with liner exchange and one-stage revisions (Table A.3 and Table A.4). Similar to the Edmonton cohort, gamma regression revealed those with Elixhauser scores of 1 or 2 and those with scores ≥3 had costs of $26,260 (p-value = 0.005), and $47,777 (p-value = 0.001), respectively, higher than those with scores of 0. Results were similar with inverse gaussian regression analysis. In addition, those >65 (compared to those <65 years) had significantly higher costs by inverse gaussian and gamma regression (Table 6).
      Table VIGamma and inverse gaussian regression adjusted costs for the combined cohort
      Gamma regressionInverse gaussian regression
      Coefficient ($)P-valueCoefficient ($)P-value
      DAIR with liner exchange (vs. DAIR)5,8590.489DAIR with liner exchange (vs. DAIR)6,6170.431
      One-stage (vs. DAIR)16,8770.318One-stage (vs. DAIR)11,8240.494
      Two-stage (vs. DAIR)45,9140.001Two-stage (vs. DAIR)51,4090.005
      Age ≥ 65 (vs <65)17,0400.014Age ≥ 65 (vs <65)19,9970.005
      Male (vs female)1,7560.796Male (vs female)3,9140.574
      Comorbidities
      Diabetes without chronic complications (vs with complications)-17,1690.126Diabetes without chronic complications (vs with complications)-21,2840.106
      Hypertension without complications (vs with complications)-21,8930.080Hypertension without complication (vs with complications)-27,7100.038
      Elixhauser Index 1 or 2 (vs. 0)26,2600.005Elixhauser Index 1 or 2 (vs. 0)30,9620.005
      Elixhauser Index ≥ 3 (vs. 0)47,7770.001Elixhauser Index ≥ 3 (vs. 0)51,2540.004
      Staphylococcus aureus (vs. non-Staphylococcus aureus)9,8060.170Staphylococcus aureus (vs. non-Staphylococcus aureus)9,5610.196

      Discussion

      This retrospective study evaluated the economic burden of surgical interventions utilized in complex SSIs following total hip and total knee arthroplasties in Edmonton, Alberta, in addition to an Edmonton and Calgary, combined cohort. These two major urban centers account for the majority, over two thirds, of all hip and knee replacements in the entire province of Alberta. This updated cohort provides a more accurate estimate of the economic burden of revision procedures for complex SSIs in Alberta by substantially increasing the number of infections studied. Costing data for just the Calgary cohort can be found in a previous publication [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. The most common procedure performed in both groups was DAIR with liner exchange; DAIR and DAIR with liner exchange were the least costly interventions. At 12 and 24-months following intervention, one-stage revision was the costliest procedure in the Edmonton cohort, while two-stage revision was the costliest for the combined cohort. The initial admission inpatient costs accounted for a majority of the accumulated 12 and 24-month costs. Those with Elixhauser comorbidity scores of one or more incurred statistically significant higher costs compared to those with scores of zero. While not meeting statistical significance, there was a trend towards increased costs in those with disease caused by Staphylococcus aureus.
      One-stage revision in the Edmonton cohort was the costliest intervention contrasting from the combined cohort in this study, in addition to the prior study which showed two-stage revision as the costliest intervention [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. Reasons for this discrepancy are multifold. First, the number of patients who underwent one-stage revision in the Edmonton group was five compared to 27 who underwent two-stage revision. The smaller sample size of the one-stage cohort reduces the precision of the estimate. Second, the higher costs may be partially explained by the comorbid status of the patients. Of the five who underwent one-stage revision in the Edmonton cohort, 4 (80%) had hypertension with complications and all (100%) had diabetes with complications. Comparatively, 70% of those undergoing the other three procedures had hypertension with complications and 86.8% - 92.5% had diabetes with complications (Table 1). Although it did not meet statistical significance, those with complications related to hypertension and diabetes amassed higher costs than those without. Thus, it is conceivable that the higher relative rates of diabetes and hypertension with complications in those undergoing one-stage revision contributed to the increased costs observed compared to the other procedures for both the Edmonton and combined cohort. LOS was also highest for one-stage revisions. As a longer LOS can correlate directly with higher per-patient in-hospital costs (compared to earlier discharge) [
      • Fine M.J.
      • Pratt H.M.
      • Obrosky D.S.
      • et al.
      Relation between length of hospital stay and costs of care for patients with community-acquired pneumonia.
      ,
      • Polverejan E.
      • Gardiner J.C.
      • Bradley C.J.
      • Holmes-Rovner M.
      • Rovner D.
      Estimating mean hospital cost as a function of length of stay and patient characteristics.
      ], the finding of one-stage revisions incurring the highest costs may also be partially explained by this. However, it has been recommended that candidates for one-stage revision be, among other factors, generally healthy and specifically without immunosuppression with the identification of a low virulence organism [
      • Ure K.J.
      • Amstutz H.C.
      • Nasser S.
      • Schmalzried T.P.
      Direct-exchange arthroplasty for the treatment of infection after total hip replacement. An average ten-year follow-up.
      ]. This group of five patients who underwent one stage revisions may not necessarily fit the typical indications suggesting an element of surgeon and patient preference into the decision-making.
      Consistent findings between this study and our prior study evaluating the Calgary zone alone, includes findings confirming that the majority of costs accrued occur within the first 12 months following the revision procedure with minimal incremental accrued costs between 12 and 24 months following the initial revision, in addition to those in the combined cohort with Elixhauser comorbidity scores of three or greater having statistically significant higher costs compared to those with scores of zero [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. These findings are congruent with pre-existing data. A temporal association between post-operative SSIs and complications, particularly in those with increased comorbidities, has been well described. Most of these adverse events typically occur within 90 days of a procedure [
      • Hartzler M.A.
      • Li K.
      • Geary M.B.
      • Odum S.M.
      • Springer B.D.
      Complications in the treatment of prosthetic joint infection.
      ,
      • Thompson J.S.
      • Baxter B.T.
      • Allison J.G.
      • et al.
      Temporal Patterns of Postoperative Complications.
      ]. As a result, healthcare utilization (e.g., emergency room visits and hospital re-admission), and associated costs, are increased within this period. Further, multimorbid patients are known to have an increased risk of post-operative complications and resource utilization [
      • Visser A.
      • Geboers B.
      • Gouma D.J.
      • Goslings J.C.
      • Ubbink D.T.
      Predictors of surgical complications: A systematic review.
      ]. Post-operative complications and increased resource utilization can lead to prolonged hospitalization which can further increase hospital associated costs [
      • Fine M.J.
      • Pratt H.M.
      • Obrosky D.S.
      • et al.
      Relation between length of hospital stay and costs of care for patients with community-acquired pneumonia.
      ,
      • Capelastegui A.
      • España P.P.
      • Quintana J.M.
      • et al.
      Improvement of process-of-care and outcomes after implementing a guideline for the management of community-acquired pneumonia: a controlled before-and-after design study.
      ]. With the passage of time, particularly after 12 months, post-operative SSIs and complications decreases resulting in decreased comparative healthcare utilization and subsequently, less accumulated cost [
      • Hartzler M.A.
      • Li K.
      • Geary M.B.
      • Odum S.M.
      • Springer B.D.
      Complications in the treatment of prosthetic joint infection.
      ,
      • Bozic K.J.
      • Stacey B.
      • Berger A.
      • Sadosky A.
      • Oster G.
      Resource utilization and costs before and after total joint arthroplasty.
      ].
      Procedures without hardware removal (DAIR and DAIR with line exchange combined) yielded lower 12-month mean costs compared to procedures with hardware removal (one-stage and two-stage revisions) due to higher inpatient costs and increased LOS. Similarly, gamma and inverse gaussian analysis in both the Edmonton and combined Edmonton and Calgary cohort revealed those undergoing two-stage revisions to have significantly higher costs when compared to those undergoing DAIR. While this information and its potential applicability should be taken in the context of a full economic evaluation, it provides insight into ways in which economic models may be constructed. Specifically, it suggests that economic models should incorporate the presence or absence of hardware removal and/or exchange in analytical methods.
      Overall, higher costs for the revision procedures were identified in this study compared to prior studies. For example, a study from Northern Finland in 2019 found the total cost of two-stage revision to be €53,413 which is equivalent to $72,350.58 CAD [
      • Puhto T.
      • Puhto A.P.
      • Vielma M.
      • Syrjälä H.
      Infection triples the cost of a primary joint arthroplasty.
      ]. Similarities to this study includes the usage of actual or micro-costing data to evaluate associated costs. However, significant differences were observed for the length of stay between the two studies. While the Finland study identified a mean LOS of 43 days for two-stage revisions, this study identified a mean LOS of 72.7 days. A similar pattern was seen with LOS comparisons for DAIR between the two studies. The increased LOS may explain the higher costs found in our study. This comparison also highlights the importance of regional factors and variations that can influence cost.
      Our study has several limitations to consider. First, although the combined cohort captures the two largest health zones in Alberta, it excludes those who obtain care outside of these regions and thus, does not fully capture the entire province. However, as a large percentage of the population (nearly 70%) reside within these two locations, in addition to serving as the tertiary referral centers for surrounding communities, these findings are likely to represent an approximation to true values for the entire population. Similar to our previous study, this study only captures those with early-onset complex SSIs (i.e., infections occurring within 3 months following initial arthroplasty) [
      • Mponponsuo K.
      • Leal J.
      • Puloski S.
      • et al.
      Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
      ]. Those with infectious complications after 3 months would not be included. In addition, unlike micro-costing, gross costing approximates but does not precisely reflect true incurred costs. This would lead to a reduction in the accuracy and precision of our findings. However, as most infectious complications occur within 90 days following arthroplasty and a majority of the total costs were driven by the inpatient costs, these two factors are unlikely to substantially alter the costs [
      • Rennert-May E.
      • Bush K.
      • Vickers D.
      • Smith S.
      Use of a provincial surveillance system to characterize postoperative surgical site infections after primary hip and knee arthroplasty in Alberta, Canada.
      ]. Next, physician-associated costs (e.g., surgical billing fees and consultations services) were not included. This typically adds an additional 10% - 15% of costs depending on the revision in question and should be ultimately accounted for in decision-making [
      • Kaye D.R.
      • Luckenbaugh A.N.
      • Oerline M.
      • et al.
      Understanding the Costs Associated With Surgical Care Delivery in the Medicare Population.
      ], particularly as those costs will be greater when a two stage procedure is utilized given that the patient will be admitted to hospital more frequently.
      Despite these limitations, there are several study strengths. While prior studies have typically focused on a single center or hospital system, our study was able to evaluate two major urban centers and their respective catchment areas, encompassing multiple hospitals, both academic and community based, and a majority of infections that occur in the single healthcare delivery system of Alberta, Canada: Alberta Health Services (AHS) [
      • Klouche S.
      • Sariali E.
      • Mamoudy P.
      Total hip arthroplasty revision due to infection: A cost analysis approach.
      ,
      • Akindolire J.
      • Morcos M.W.
      • Marsh J.D.
      • et al.
      The economic impact of periprosthetic infection in total hip arthroplasty.
      ]. By combining the Edmonton and Calgary data, this study increased the sample size and provided a more comprehensive understanding of the costs and a more accurate assessment of the economic impact of the surgical management of complex SSIs of the knee and hip within AHS. Our ability to utilize micro-costing for both urban centers, which is not available in many jurisdictions, for inpatient resource utilization also improves the accuracy of the final estimates, particularly as most costs were accrued within the inpatient context. This information would be applicable to many other centers both nationally and internationally. The study also confirms findings from our previous study that revealed a majority of the total incurred costs post-operatively, accumulated within the first 12-months following the revision procedure, presumably due to peri-operative complications. This confirms the need for policies and interventions throughout the entire healthcare delivery system that are targeted at reducing and/or mitigating post-operative complications and their costs such as streamlined discharge services with dedicated teams. Finally, by combining the economic data with clinical outcomes, policy-makers and clinicians can further make informed decisions that optimize patient outcomes and cost.

      Conclusion

      Complex SSI following hip and knee prosthesis are uncommon but serious complications with significant attributable costs. This study revealed DAIR followed by DAIR with liner exchange resulted in the lowest 12 and 24 month cost following the initial revision procedure in Alberta, Canada. Most of the costs, regardless of the procedure, were acquired in the first 12 months. Finally, those with Elixhauser comorbidity scores of >3 accrued significantly higher costs compared to those with scores of zero. These findings are necessary to understand the economic burden of managing SSIs following hip and knee replacements, stratified by surgical management type. This information makes suggestions around which patients and surgeries are most costly to the healthcare system which can be used to plan further studies on interventions that might target high resource utilization. Ultimately, these findings should and will be combined with clinical outcomes data and the effectiveness of each surgical strategy at preventing further infection and other complications, in order to complete a full economic evaluation and cost-effectiveness study. Once it is fully understood which management strategies offer the best value for health outcomes for different patients with complex SSIs following hip and knee arthroplasties, clinical guidelines and health policy can be adjusted in order to optimize costs and patient outcomes.

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Declaration of interest

      None

      Conflicts of interest

      The authors have no conflicts of interest to declare.

      Acknowledgements

      The authors would like to acknowledge the Alberta Health Services Infection Prevention and Control (IPC) group for in providing us the cohort of patients with SSIs.

      APPENDIX.

      Table A.1Subgroup analysis of mean cost, number of hospitalizations, and length of stay 12 months postoperatively for the Edmonton cohort.
      All patientsDAIRDAIR with liner exchangeOne-stageTwo-stageNo Surgical management
      52,124

      (37,040 – 67,208)
      62,212

      (50,749 – 73,675)
      148,101

      (41,347 – 33,755)
      111,858

      (77,515 – 146,201)
      26,846

      (7,361 – 46,330)
      Age <65
      N26673162
      Hospital admissions (N)10.7

      (6.3 – 15.1)
      7.9

      (5.8 – 10.0)
      18.0

      (0.6 – 35.4)
      20.8

      (13.8 – 27.8)
      6.5

      (1.7 – 14.7)
      LOS (days)26.0

      (12.0 – 40.0)
      23.3

      (16.4 – 30.1)
      86.3

      (23.6 – 196.2)
      70.8

      (34.3 – 107.2)
      17.5

      (1.9 – 33.1)
      Cost ($)53,216

      (29,824 – 76,608)
      45,964

      (36,770 – 55,157)
      125,349

      (9,738.8 – 260,436)
      119,076

      (71,729 – 166,424)
      24,941

      (13,957 – 39,926)
      Age >65
      N27822123
      Hospital admissions (N)7.4

      (4.9 – 9.8)
      12.8

      (9.4 – 16.1)
      16.0

      (5.5 – 37.5)
      17.8

      (3.1 – 32.4)
      12.0

      (1.3 – 25.3)
      LOS (days)33.8

      (15.5 – 52.0)
      49.4

      (31.6 – 67.3)
      113.0

      (76.8 – 302.8)
      67.8

      (19.6 – 115.9)
      19.7

      (9.2 – 30.1)
      Cost ($)51,072

      (30,280 – 71,864)
      75,488

      (56,324 – 94,650)
      182,229

      (64,577 – 429,036)
      102,235

      (44,311 – 160,158)
      28,115

      (9,833 – 46,397)
      Elixhauser Comorbidities
      0
      N16682122
      Hospital admissions (N)7.1

      (4.4 – 9.8)
      7.9

      (5.5 – 10.4)
      9.0

      (2.0 – 17.0)
      14.6

      (8.2 – 21.1)
      4.5

      (1.4 – 7.6)
      LOS (days)23.9

      (12.6 – 35.2)
      32.3

      (12.7 – 51.9)
      8.5

      (4.5 – 12.5)
      51.3

      (7.7 – 94.8)
      11.0

      (6.4 – 28.4)
      Cost ($)48,870

      (28,944 – 68,797)
      51,956

      (33,757 – 70,154)
      30,734

      (26,878 – 34,591)
      90,254

      (31,448– 149,061)
      18,557

      (672 – 36,442)
      1 or 2
      N2348261
      Hospital Admissions (N)10.4

      (5.5 – 15.4)
      13.5

      (9.0 – 18.1)
      24.0

      (7.9 – 40.1)
      26.3

      (10.1 – 42.6)
      10.0
      LOS (days)38.3

      (14.0 – 62.7)
      40.1

      (27.4 – 52.9)
      126.0

      (7 – 245)
      97.8

      (19.1 – 176.6)
      28.0
      Cost ($)60,862

      (28,947 – 92,778)
      64,013

      (48,907 – 79,118)
      174,046

      (19,338 – 347,430)
      148,448

      (48,286 – 248,610)
      32,733
      ≥3
      N14331102
      Hospital admissions (N)8.9

      (4.5 – 13.2)
      11.7

      (7.3 – 16.1)
      20.021.2

      (3.5 – 38.9)
      15.0

      (4.2 – 25.8)
      LOS (days)23.1

      (10.1 – 36.1)
      45.2

      (27.7 – 62.7)
      216.074.3

      (24.0 – 124.6)
      22.0

      (8.2 – 52.2)
      Cost ($)41,486

      (24,377 – 58,595)
      80,726

      (51,500 – 109,952)
      330,946115,829

      (58,291 – 173,368)
      32,190

      (8,734 – 73,115)
      Table A.2Subgroup analysis of mean cost, number of hospitalizations, and length of stay 24 months postoperatively for the Edmonton cohort.
      All patients

      ($)
      DAIRDAIR with liner exchangeOne-stageTwo-stageNo Surgical management
      52,713

      (37,202 – 68,225)
      62,553

      (51,043 – 74,064)
      68,234

      (41,347 – 337,549)
      113,701

      (79,109 – 148,293)
      26,981

      (7,633 – 46,328)
      Age <65
      N26673162
      Hospital admissions (N)10.7

      (6.3 – 15.1)
      8.0

      (5.8 – 10.1)
      18.0

      (0.5 – 35.5)
      21.5

      (13.7 – 29.3)
      6.5

      (1.8 – 11.3)
      LOS (days)26.0

      (12.0 – 40.0)
      23.3

      (16.4 – 30.3)
      86.3

      (8.9 – 164.0)
      72.1

      (35.5 – 108.6)
      17.5

      (6.6 – 28.4)
      Cost ($)53,494

      (29,915 – 77,074)
      46,287

      (36,902 – 55,671)
      125,349

      (21,267 – 229,430)
      122,109

      (74,256 – 169,963)
      25,017

      (425 – 49,609)
      Age >65
      N27822123
      Hospital admissions (N)7.6

      (4.8 – 10.5)
      12.8

      (9.5 – 16.2)
      16.0

      (5.6 – 26.4)
      17.8

      (3.1 – 32.4)
      12.0

      (1.4 – 22.6)
      LOS (days)34.2

      (15.2 – 53.2)
      49.5

      (31.7 – 67.4)
      113.0

      (58.9 – 167.1)
      67.7

      (19.6 – 115.9)
      19.7

      (9.4 – 48.8)
      Cost ($)51,962

      (30,094 – 73,829)
      75,845

      (56,648 – 95,041)
      182,229

      (48,096 – 316,362)
      102,490

      (44,617 – 160,363)
      28,289

      (906 – 55,672)
      Elixhauser Comorbidities
      0
      N16682122
      Hospital admissions (N)7.1

      (4.4 – 9.8)
      8.0

      (5.5 – 10.5)
      9.0

      (2.0 – 16.0)
      14.7

      (8.2 – 21.1)
      4.5

      (1.4 – 7.6)
      LOS (days)23.9

      (12.6 – 35.2)
      32.4

      (12.7 – 52.0)
      8.5

      (5.1 – 11.9)
      51.3

      (7.7 – 94.8)
      11.0

      (6.4 – 15.6)
      Cost ($)49,290

      (29,049 – 69,532)
      52,244

      (33,948 – 70,540)
      30,734

      (3,203 – 58,265)
      90,843

      (32,146 – 149,540)
      18,764

      (183 – 37,344)
      1 or 2
      N2348261
      Hospital admissions (N)10.8

      (5.5 – 15.9)
      13.6

      (9.0 – 18.1)
      24.0

      (8.2 – 39.8)
      28.2

      (8.9 – 47.4)
      10.0
      LOS (days)38.8

      (13.7 – 64.0)
      40.1

      (27.4 – 52.9)
      126.0

      (53.8 – 198.2)
      101.3

      (23.5 – 179.2)
      28.0
      Cost ($)61,889

      (28,936 – 94,841)
      64,068

      (48,968 – 79,168)
      174,046

      (17,823 – 330,269)
      155,545

      (54,768 – 256,323)
      32,733
      ≥3
      N14331102
      Hospital admissions (N)8.9

      (4.5 – 13.2)
      11.8

      (7.4 – 16.3)
      20.021.2

      (3.4 – 38.9)
      15.0

      (4.3 – 25.7)
      LOS (days)23.1

      (10.1 – 36.1)
      45.4

      (27.8 – 63.0)
      216.074.3

      (24.1 – 124.6)
      22.0

      (8.2 – 35.8)
      Cost41,552

      (24,440 – 58,663)
      81,595

      (52,261 – 110,928)
      330,946116,025

      (58,570 – 173,479)
      32,321

      (8,730 – 55,912)
      Table A.3Subgroup analysis of mean cost, number of hospitalizations, and length of stay 12 months postoperatively for the combined cohort.
      All patientsDAIRDAIR with Liner ExchangeOne-stageTwo-stageNo Surgical Management
      53,197

      (38,006 – 68,388)
      57,340

      (48,576 – 66,105)
      76,954

      (31,299 – 122,609)
      114,030

      (83,959 – 144,100)
      23,146

      (5,612 – 40,479)
      Age <65
      N351055203
      Hospital admissions (N)9.0

      (5.6 – 12.4)
      7.2

      (5.7 – 8.7)
      13.2

      (3.3 – 23.1)
      20.1

      (14.1 – 26.0)
      4.3

      (0.3 – 9.0)
      LOS (days)23.9

      (12.7 – 35.1)
      21.1

      (15.2 – 26.9)
      58.0

      (7.6 – 108.4)
      63.9

      (34.4 – 93.2)
      11.7

      (0.9 – 25.2)
      Cost ($)58,479

      (29,211 – 87,748)
      41,308

      (34,094 – 48,522)
      93,383

      (31,463 – 155,303)
      112,656

      (74,708 – 150,603)
      18,176

      (2,616 – 33,735)
      Age >65
      N4313913163
      Hospital admissions (N)8.1

      (5.9 – 10.3)
      11.5

      (9.2 – 13.8)
      9.5

      (2.8 – 16.1)
      22.4

      (6.0 – 38.9)
      12.0

      (1.3 – 22.7)
      LOS (days)32.0

      (18.9 – 45.2)
      43.2

      (31.3 – 55.1)
      38.8

      (1.0 – 76.6)
      83.7

      (36.6 – 130.7)
      19.7

      (5.2 – 34.2)
      Cost ($)48,898

      (33,811 – 63,984)
      69,451

      (55,303 – 83,599)
      70,635

      (18,674 – 122,596)
      115,747

      (62,204 – 169,290)
      28,115

      (213 – 56,017)
      Elixhauser Comorbidities
      0
      N2911513133
      Hospital admissions (N)7.1

      (4.6 – 9.7)
      7.6

      (5.9 – 9.3)
      6.1

      (3.2 – 9.3)
      14.5

      (8.6 – 20.4)
      3.0

      (0.4 – 5.6)
      LOS (days)25.5

      (13.6 – 37.4)
      26.8

      (14.7 – 39.0)
      13.6

      (9.0 – 18.2)
      50.1

      (10.3 – 89.8)
      7.3

      (2.0 – 12.6)
      Cost ($)47,330

      (31,142 – 63,518)
      45,925

      (34,405 – 57,444)
      39,645

      (31,664 – 47,625)
      91,631

      (37,997 – 145,264)
      13,920

      (1,913 – 25,927)
      1 or 2
      N2979391
      Hospital Admissions (N)9.2

      (5.2 – 13.2)
      11.8

      (8.5 – 15.0)
      16.7

      (1.6 – 31.8)
      21.2

      (10.0 – 32.5)
      10.0
      LOS (days)32.3

      (12.8 – 51.9)
      36.8

      (25.6 – 47.9)
      90.0

      (19.2 – 160.8)
      80.0

      (27.8 – 132.2)
      28.0
      Cost ($)53,384

      (27,777 – 78,991)
      60,450

      (45,458 – 75,443)
      126,374

      (2,687 – 251,184)
      124,035

      (58,684 – 189,385)
      32,733
      ≥3
      N20502142
      Hospital admissions (N)9.5

      (6.0 – 13.0)
      10.9

      (7.5 – 14.3)
      30.0

      (9.2 – 50.8)
      27.6

      (8.6 – 45.9)
      15.0

      (3.7 – 26.3)
      LOS (days)26.9

      (14.9 – 38.8)
      44.6

      (30.4 – 58.7)
      173.5

      (69.4 – 277.6)
      88.9

      (39.3 – 138.5)
      22.0

      (8.6 – 35.4)
      Cost ($)61,433

      (17,221 – 105,646)
      78,683

      (54,921 – 102,45)
      245,337

      (54,324 – 436,350)
      128,397

      (74,380 – 182,413)
      32,190

      (10,094 – 54,586)
      Table A.4Subgroup analysis of mean cost, mean number of hospitalizations, and mean length of stay 24 months postoperatively for the combined cohort.
      All patients ($)DAIRDAIR with liner exchangeOne-stageTwo-stageNo Surgical management
      53,808

      (40,879 – 66,737)
      57,789

      (50,429 – 65,150)
      77,250

      (39,633 – 114,867)
      116,170

      (91,225 – 141,115)
      23,324

      (9,695 – 36,953)
      Age <65
      N351055203
      Hospital admissions (N)9.0

      (5.6 – 12.4)
      7.3

      (5.8 – 8.9)
      14.2

      (3.5 – 24.9)
      20.6

      (14.0 – 27.2)
      4.3

      (0.3 – 9.0)
      LOS (days)23.9

      (12.7 – 35.1)
      21.3

      (15.4 – 22.3)
      58.2

      (7.6 – 108.8)
      64.9

      (35.3 – 94.5)
      11.7

      (1.9 – 21.5)
      Cost ($)58,701

      (29,359 – 88,044)
      41,872

      (34,556 – 49,189)
      94,072

      (31,614 – 156,529)
      115,151

      (76,722 – 153,580)
      18,358

      (2,622 – 34,094)
      Age >65
      N4313913163
      Hospital admissions (N)8.4

      (6.0 – 10.8)
      11.5

      (9,2 – 13.8)
      9.5

      (2.8 – 16.1)
      22.8

      (6.5 – 39.2)
      12.0

      (1.3 – 22.7)
      LOS (days)33.5

      (31.4 – 55.2)
      43.3

      (31.4 – 55.2)
      38.8

      (1.0 – 76.6)
      84.5

      (38.9 – 131.1)
      19.7

      (5.0 – 33.7)
      Cost ($)49,825

      (33,890 – 65,759)
      69,812

      (55,659 – 83,966)
      70,779

      (18,805 – 122,755)
      117,443

      (64,782 – 170,105)
      28,115

      (22,599 – 33,980)
      Elixhauser Comorbidities
      0
      N2911513133
      Hospital admissions (N)7.3

      (4.6 – 10.0)
      7.7

      (6.0 – 9.4)
      6.5

      (3.2 – 9.7)
      14.5

      (8.6 – 20.4)
      3.0

      (0.4 – 5.6)
      LOS (days)27.3

      (12.5 – 42.1)
      26.9

      (14.7 – 39.1)
      13.7

      (9.0 – 18.3)
      50.1

      (10.3 – 89.8)
      7.3

      (2.6 – 12.0)
      Cost ($)48,104

      (31,183 – 65,026)
      46,335

      (34,765 – 57,904)
      40,000

      (31,974 – 48,027)
      92,207

      (38,675 – 145,739)
      14,189

      (1,935 – 26,443)
      1 or 2
      N2979391
      Hospital admissions (N)9.4

      (5.3 – 13.6)
      11.9

      (8.6 – 15.1)
      16.7

      (1.6
      22.4

      (9.4 – 35.5)
      10.0
      LOS (days)32.7

      (12.6 – 52.9)
      36.9

      (25.8 – 48.1)
      90.0

      (18.3 -
      82.3

      (30.0 – 134.6)
      28.0
      Cost ($)54,220

      (27,796 – 80,645)
      60,824

      (45,861 – 75,788)
      126,373

      (9,496 – 243,251)
      128,790

      (61,919 – 195,661)
      32,733
      ≥3
      N20502142
      Hospital admissions (N)9.5

      (6.0 – 13.0)
      11.0

      (7.6 – 14.4)
      30.0

      (6.8 – 53.2)
      27.6

      (9.2 – 46.1)
      15.0

      (3.7 -
      LOS (days)26.9

      (14.9 – 38.8)
      44.7

      (30.5 – 58.9)
      173.5

      (55.7 – 291.3)
      89.9

      (40.8 – 138.9)
      22.0

      (7.7 -
      Cost61,480

      (17,272 – 105,688)
      79,338

      (55,529 – 103,147)
      245,684

      (72,096 – 419,272)
      130,308

      (77,624 – 189,993)
      32,321

      (9,485 – 55,157)

      Bibliography

      1. Canadian Institute for Health Information. Hip and Knee Replacements in Canada, 2017–2018: Canadian Joint Replacement Registry Annual Report. Ottawa, ON: CIHI; 2019.

      2. Arthritis Society. (2021). Facts and Figures. Accessed from: https://arthritis.ca/about-arthritis/what-isarthritis/arthritis-facts-and-figures

        • Miletic K.G.
        • Taylor T.N.
        • Martin ET Vaidya R.
        • Kaye K.S.
        Readmissions after diagnosis of surgical site infection following knee and hip arthroplasty.
        Infect Control Hosp Epidemiol. 2014; 35: 152-157
        • Nair R.
        • Schweizer M.L.
        • Singh N.
        Septic Arthritis and Prosthetic Joint Infections in Older Adults.
        Infect Dis Clin North Am. 2017; 31: 715-729
        • Triffault-Fillit C.
        • Ferry T.
        • Laurent F.
        • et al.
        Microbiologic epidemiology depending on time to occurrence of prosthetic joint infection: a prospective cohort study.
        Clin Microbiol Infect. 2019; 25: 353-358
      3. American Academy of Orthopaedic Surgeons EvidenceBased Clinical Practice Guideline for Diagnosis and Prevention of Periprosthetic Joint Infections. https://www.aaos.org/pjiguideline Published March 11, 2019

        • Lange J.
        • Troelsen A.
        • Thomsen R.W.
        • Søballe K.
        Chronic infections in hip arthroplasties: comparing risk of reinfection following one-stage and two-stage revision: a systematic review and meta-analysis.
        Clin Epidemiol. 2012; 4: 57-73
        • Silvestre A.
        • Almeida F.
        • Renovell P.
        • Morante E.
        • López R.
        Revision of infected total knee arthroplasty: two-stage reimplantation using an antibiotic-impregnated static spacer.
        Clin Orthop Surg. 2013; 5: 180-187
        • Kunutsor S.K.
        • Whitehouse M.R.
        • Blom A.W.
        • Beswick A.D.
        Re-Infection Outcomes following One- and Two-Stage Surgical Revision of Infected Hip Prosthesis: A Systematic Review and Meta-Analysis.
        PLoS One. 2015; 10e0139166
        • Tsang S.J.
        • Ting J.
        • Simpson A.
        • Gaston P.
        Outcomes following debridement, antibiotics and implant retention in the management of periprosthetic infections of the hip: a review of cohort studies.
        Bone Joint J. 2017; 99-b: 1458-1466
        • Qasim S.N.
        • Swann A.
        • Ashford R.
        The DAIR (debridement, antibiotics and implant retention) procedure for infected total knee replacement - a literature review.
        Sicot j. 2017; 3: 2
        • Mponponsuo K.
        • Leal J.
        • Puloski S.
        • et al.
        Economic burden of surgical management of surgical site infections following hip and knee replacements in Calgary, Alberta, Canada.
        Infection Control & Hospital Epidemiology. 2021; : 1-8
      4. Government of Alberta. (2021). Population Projects: Alberta and Census Divisions, 2021–2046. https://www.alberta.ca/population-statistics.aspx#data.

        • Rennert-May E.D.
        • Conly J.
        • Smith S.
        • et al.
        The cost of managing complex surgical site infections following primary hip and knee arthroplasty: A population-based cohort study in Alberta, Canada.
        Infect Control Hosp Epidemiol. 2018; 39: 1183-1188
        • Clement Nee Shrive F.M.
        • Ghali W.A.
        • Donaldson C.
        • Manns B.J.
        The impact of using different costing methods on the results of an economic evaluation of cardiac care: microcosting vs gross-costing approaches.
        Health Econ. 2009; 18: 377-388
      5. Canadian Institute for Health Information. DAD resource intensity weights and expected length of stay (ELOS) for CMG+ 2020, https://secure.cihi.ca/estore/productSeries.htm?pc=PCC90; 2020 [accessed 20 January 2022].

      6. Statistics Canada. Consumer Price Index: Annual Review, 2019. https://www150.statcan.gc.ca/n1/daily-quotidien/200122/dq200122c-eng.htm. Accessed November 2, 2020.

        • Fine M.J.
        • Pratt H.M.
        • Obrosky D.S.
        • et al.
        Relation between length of hospital stay and costs of care for patients with community-acquired pneumonia.
        Am J Med. 2000; 109: 378-385
        • Polverejan E.
        • Gardiner J.C.
        • Bradley C.J.
        • Holmes-Rovner M.
        • Rovner D.
        Estimating mean hospital cost as a function of length of stay and patient characteristics.
        Health Economics. 2003; 12: 935-947
        • Ure K.J.
        • Amstutz H.C.
        • Nasser S.
        • Schmalzried T.P.
        Direct-exchange arthroplasty for the treatment of infection after total hip replacement. An average ten-year follow-up.
        J Bone Joint Surg Am. 1998; 80: 961-968
        • Hartzler M.A.
        • Li K.
        • Geary M.B.
        • Odum S.M.
        • Springer B.D.
        Complications in the treatment of prosthetic joint infection.
        The Bone & Joint Journal. 2020; 102-B: 145-150
        • Thompson J.S.
        • Baxter B.T.
        • Allison J.G.
        • et al.
        Temporal Patterns of Postoperative Complications.
        Archives of Surgery. 2003; 138: 596-603
        • Visser A.
        • Geboers B.
        • Gouma D.J.
        • Goslings J.C.
        • Ubbink D.T.
        Predictors of surgical complications: A systematic review.
        Surgery. 2015; 158: 58-65
        • Capelastegui A.
        • España P.P.
        • Quintana J.M.
        • et al.
        Improvement of process-of-care and outcomes after implementing a guideline for the management of community-acquired pneumonia: a controlled before-and-after design study.
        Clin Infect Dis. 2004; 39: 955-963
        • Bozic K.J.
        • Stacey B.
        • Berger A.
        • Sadosky A.
        • Oster G.
        Resource utilization and costs before and after total joint arthroplasty.
        BMC Health Serv Res. 2012; 12: 73
        • Puhto T.
        • Puhto A.P.
        • Vielma M.
        • Syrjälä H.
        Infection triples the cost of a primary joint arthroplasty.
        Infect Dis (Lond). 2019; 51: 348-355
        • Rennert-May E.
        • Bush K.
        • Vickers D.
        • Smith S.
        Use of a provincial surveillance system to characterize postoperative surgical site infections after primary hip and knee arthroplasty in Alberta, Canada.
        Am J Infect Control. 2016; 44: 1310-1314
        • Kaye D.R.
        • Luckenbaugh A.N.
        • Oerline M.
        • et al.
        Understanding the Costs Associated With Surgical Care Delivery in the Medicare Population.
        Ann Surg. 2020; 271: 23-28
        • Klouche S.
        • Sariali E.
        • Mamoudy P.
        Total hip arthroplasty revision due to infection: A cost analysis approach.
        Orthopaedics & Traumatology: Surgery & Research. 2010; 96: 124-132
        • Akindolire J.
        • Morcos M.W.
        • Marsh J.D.
        • et al.
        The economic impact of periprosthetic infection in total hip arthroplasty.
        Can J Surg. 2020; 63: E52-e56