Socioeconomic burden of total joint arthroplasty for symptomatic hip and knee osteoarthritis in the Italian population: A 5-year analysis based on hospitalization records

Authors

  • P. Piscitelli,

    Corresponding author
    1. University of Florence, Florence, and ISBEM Research Centre, Brindisi, Italy
    • Department of Internal Medicine, Largo Palagi 1, 50138 Florence, Italy
    Search for more papers by this author
    • Dr. Piscitelli has received consultant fees and/or speaking fees (less than $10,000) from Amgen.

  • G. Iolascon,

    1. Second University of Naples, Naples, Italy
    Search for more papers by this author
    • Drs. Iolascon, Distante, Gimigliano, Brandi, and Migliore have received consultant fees and/or speaking fees (less than $10,000 each) from Merck, Chiesi, Glaxo, Sanofi-Aventis, Novartis, Roche, Stroder-Servier, Ely Lilly, and Nycomed.

  • G. Di Tanna,

    1. University of Rome La Sapienza, Rome, Italy
    Search for more papers by this author
    • Dr. Di Tanna has received consultant fees and/or speaking fees (less than $10,000) from Novartis.

  • E. Bizzi,

    1. Fatebenefratelli St. Peter's Hospital, Rome, Italy
    Search for more papers by this author
  • G. Chitano,

    1. ISBEM Research Centre, Brindisi, Italy
    Search for more papers by this author
  • A. Argentiero,

    1. ISBEM Research Centre, Brindisi, and University of Pisa, Pisa, Italy
    Search for more papers by this author
  • C. Neglia,

    1. ISBEM Research Centre, Brindisi, Italy
    Search for more papers by this author
  • L. Giolli,

    1. ISBEM Research Centre, Brindisi, Italy
    Search for more papers by this author
  • A. Distante,

    1. ISBEM Research Centre, Brindisi, and University of Pisa, Pisa, Italy
    Search for more papers by this author
    • Drs. Iolascon, Distante, Gimigliano, Brandi, and Migliore have received consultant fees and/or speaking fees (less than $10,000 each) from Merck, Chiesi, Glaxo, Sanofi-Aventis, Novartis, Roche, Stroder-Servier, Ely Lilly, and Nycomed.

  • R. Gimigliano,

    1. Second University of Naples, Naples, and Casa di Cura Santa Maria del Pozzo, Somma Vesuviana, Italy
    Search for more papers by this author
    • Drs. Iolascon, Distante, Gimigliano, Brandi, and Migliore have received consultant fees and/or speaking fees (less than $10,000 each) from Merck, Chiesi, Glaxo, Sanofi-Aventis, Novartis, Roche, Stroder-Servier, Ely Lilly, and Nycomed.

  • M. L. Brandi,

    1. University of Florence, Florence, Italy
    Search for more papers by this author
    • Drs. Iolascon, Distante, Gimigliano, Brandi, and Migliore have received consultant fees and/or speaking fees (less than $10,000 each) from Merck, Chiesi, Glaxo, Sanofi-Aventis, Novartis, Roche, Stroder-Servier, Ely Lilly, and Nycomed.

  • Alberto Migliore

    1. Fatebenefratelli St. Peter's Hospital, Rome, Italy
    Search for more papers by this author
    • Drs. Iolascon, Distante, Gimigliano, Brandi, and Migliore have received consultant fees and/or speaking fees (less than $10,000 each) from Merck, Chiesi, Glaxo, Sanofi-Aventis, Novartis, Roche, Stroder-Servier, Ely Lilly, and Nycomed.


Abstract

Objective

To assess the burden of total joint arthroplasties (TJAs) performed for symptomatic hip and knee osteoarthritis (OA) in the Italian population.

Methods

We analyzed national hospitalizations and diagnosis-related group databases to compute incidence, annual percent change (APC), direct costs, and working days lost between 2001 and 2005 following TJA due to OA.

Results

In 2005, we recorded a total of 41,816 (APC +5.4; 95% confidence interval [95% CI] 5.1–5.8) and 44,051 (APC +13.4; 95% CI 13.1–13.8) hip and knee arthroplasties, respectively. Women represented the majority of patients undergoing TJA procedures (female:male ratio 1.7:1 for hip arthroplasties and 2.9:1 for knee arthroplasties). When analyzing the data by age groups, most of the patients were in the age groups 65–74 years and ≥75 years, although the highest increases were observed in those ages <65 years. Revisions accounted for 6,387 (APC +4.9; 95% CI 4.0–5.7) and 2,295 (APC +17.4; 95% CI 15.7–19.2) procedures for the hip and knee, respectively. Loss of working days in patients ages <65 years was estimated between 805,000 and 1 million days. Hospital costs increased from 741 million to 1 billion euros over the 5-year period (from 412 to 538 million euros for hip arthroplasties and from 329 to 517 million euros for knee arthroplasties). Rehabilitation costs increased from 228 to 322 million euros. Postoperative complications were estimated between 3.1 and 4.4 million euros. The average costs per patient were 16,835 and 15,358 euros for hip and knee arthroplasties, respectively.

Conclusion

The socioeconomic burden of TJAs performed for symptomatic OA in Italy is remarkable and calls for the adoption of proper preventive measures.

INTRODUCTION

Osteoarthritis (OA) is an evolving disease and a major cause of impaired mobility resulting in marked reduction of quality of life and relevant costs (1, 2). The hip and knee are the joints most frequently affected by OA and are associated with moderate to severe disability even in younger adults (3). Furthermore, this condition may result in lower extremity disability and impair independence and psychological functioning of the affected patients, also leading to relevant socioeconomic consequences and financial loss (4). According to the World Health Organization, OA is the sixth-leading cause of disability in the world (5), being comparable to that of asthma (6). The prevalence of OA increases indefinitely with age because the condition is not spontaneously reversible (7). Almost 9.6% of men and 18.0% of women ages ≥60 years in the world are thought to have symptomatic OA (1). Men are affected more often than women ages <45 years, whereas women are affected more frequently after age 55 years (1). Limitations to job activities are relevant in people with OA if compared with a healthy age- and sex-matched population, thus causing a reduction of working hours, problems in applying for jobs, or early retirement due to the illness (8). Given the increasing incidence of OA with age, the extended life expectancy observed in Italy is expected to result in a progressively higher number of people with this condition. Currently, 20% of the Italian population is age >65 years (9); therefore, Italy may represent an interesting case study, anticipating possible scenarios occurring in other European countries concerning the burden of chronic degenerative diseases. The diagnosis of OA already accounts for 55% of patients admitted to rheumatology units (10), and the expected medical costs of OA patients are double those of subjects without OA (11). These costs are related to arthroplasty and nonsteroidal antiinflammatory drug (NSAID) consumption, as well as other drugs commonly used to prevent their gastrointestinal side effects. Rehabilitation is also a relevant and expensive part of the treatment. The incidence of total hip replacement in Europe varies between 50 and 140 procedures per 100,000 inhabitants (12), with OA being the main cause for intervention. According to UK data, the average cost of hip replacement exceeds £4,000 (€6,500), but there is a lack of specific data concerning Italy and other European Union countries (12).

The aim of this work was to evaluate the socioeconomic burden of total joint arthroplasties (TJAs) performed in Italy for symptomatic hip and knee OA through the analysis of national hospitalization records.

Significance & Innovations

  • The number of total hip and knee replacements due to osteoarthritis has increased during the study period (2001–2005).

  • The number of knee arthroplasties has reached that of hip arthroplasties in 2004. The annual increase versus the previous year (annual percent increase [APC]) is substantial for hip arthroplasty (APC +5.4) and impressive for knee arthroplasty (APC +13.4).

  • Despite accounting for a minor absolute number of prostheses implanted, the highest increase was observed in men for both hip and knee arthroplasties.

  • The majority of revision interventions consist in hip prosthesis revisions (ratio of hip:knee revision 3:1), but knee revisions show a higher increase (APC +17.4) compared to that of the hip (APC +4.9).

MATERIALS AND METHODS

Information concerning all hospital admissions at Italian hospitals is maintained by the Italian Ministry of Health (national hospitalization database; SDO). This information is anonymous and includes patient age, diagnosis, procedures performed, and length of stay (LOS). Thanks to the availability of this huge database, we searched for the number of hip and knee arthroplasties performed in the Italian population due to the main diagnosis of OA. We also assessed costs concerning hospitalization, rehabilitation, and complications. For patients ages <65 years, loss of productivity in terms of working days lost was estimated. Hip replacements performed in patients ages <25 years were excluded from the analysis because they were considered as unlikely to be caused by OA. We selected the years 2001–2005 because it was the most recent time period available for our inquiry. Population data were obtained from the National Institute for Statistics for each of the considered years (9). Total hip arthroplasties (THAs) were defined by the following International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for surgical procedures: 81.51 and 81.53 (revision surgery). Because not all of these interventions are carried out due to a main diagnosis of OA, given that hip fractures, osteonecrosis, and tumors often result in hip arthroplasties, we performed a specific analysis using the region of Tuscany database, a region with an aging index comparable to the national average value, in order to determine the rate of hip arthroplasties attributable to a major diagnosis of OA. A diagnosis of OA was defined by ICD-9-CM code 715. The analysis of the Tuscany regional health care system databases showed that approximately 75% of all THAs were due to a main diagnosis of OA. Based on this analysis, 75% of hip arthroplasties performed all over Italy were considered as being associated with a main diagnosis of hip OA. Total knee replacements were performed mostly because of OA (this diagnosis accounted for more than 95% of interventions in the regional databases), and they were defined by the ICD-9-CM codes 81.54 and 81.55 (revision surgery) for surgical procedures. Data were stratified by sex and by age into 4 age groups (25–44, 45–64, 65–74, and ≥75 years). Using results from the analysis of the Tuscany databases, the mean incidence of postoperative complications was assumed not to exceed 0.5%. This assumption was consistent with the lowest rates reported in the current medical literature for each considered variable: postoperative infections (13–19), pulmonary embolism (19–23), deep vein thrombosis (24–27), and mortality (19, 28–35). We estimated the loss of productivity among patients ages <65 years in terms of working days lost by analyzing the average hospitalization LOS and rehabilitation period, which resulted from the analysis of the national hospitalization database.

Analyses of direct costs were based on the costs ascribed to diagnosis-related groups (DRGs), according to Ministerial Decree 549 (June 30, 1997). This law defined the national DRGs rate list adopted during the 6 years examined in our study, the values of which have been revised after 2005. The DRG considered for the cost analysis of both hip and knee TJA procedures performed because of OA was DRG 209, defined as “surgical procedures on major joints and hip replacement,” which corresponded to a fixed value of 7,979.78 euros per patient. On the basis of Tuscan databases, we computed that 75% of DRG 209 was attributable to TJA procedures (including hip and knee arthroplasties) performed because of OA, with the remaining 25% of DRG 209 being associated with the main diagnosis of hip fracture. Within DRG 209, the specific weight of hip and knee arthroplasties was computed based on the number of procedures performed each year. Cost analysis for the evaluation of complications following surgical interventions was based on the following DRGs: DRG 418 (postoperative infections), DRG 128 (deep vein thrombosis), and DRG 78 (pulmonary embolism), corresponding to costs of €2,776, €2,630, and €4,297, respectively. Based on specific estimations provided by the Italian Society for Physical and Rehabilitative Medicine, we assumed that 20% of patients undergo in-hospital rehabilitation in order to assess the burden of rehabilitation following TJA. The official Italian rate list for in-hospital rehabilitation has fixed a cost of 256 euros per day over a period of 20 days. Therefore, each patient undergoing in-hospital rehabilitation generates a cost of 5,120 euros. Furthermore, we had to consider that in Italy all patients are treated at home by physiotherapists on behalf of the local health authorities. According to Italian national rate lists, rehabilitation carried out in patients' homes has an average cost of 80 euros per day and is continued for a minimum period of 30 days, resulting in a cost of 2,400 euros per patient, although some patients may carry on up to 60 and even 90 days.

RESULTS

In 2005 we recorded a total of 41,816 hip arthroplasties (Table 1) and 44,051 knee arthroplasties (Table 2) due to OA. For hip arthroplasties, we computed an overall annual percent change (APC) across the 5 examined years of +5.4 (95% confidence interval [95% CI] 5.1–5.8), with the APC for men being +7.0 (95% CI 6.4–7.5) and the APC for women being +4.5 (95% CI 4.1–4.9). For knee arthroplasties, we computed an overall APC across the 5 years examined of +13.4 (95% CI 13.1–13.8), with the APC for men being +16.6 (95% CI 15.8–17.3) and the APC for women being +12.4 (95% CI 12.0–12.8).

Table 1. Number of hip arthroplasties performed in Italy between 2001 and 2005 and APC shown by sex and age groups*
 20012002200320042005APC (95% CI)
  • *

    APC = annual percentage change; 95% CI = 95% confidence interval.

Age group, years      
 25–44      
  Men5836386587918329.8 (7.2–12.4)
  Women6146826767467775.7 (3.3–8.3)
 45–64      
  Men3,9884,1574,4155,0315,0666.9 (5.9–7.9)
  Women5,1125,3395,3975,9165,8413.8 (2.9–4.6)
 65–74      
  Men5,0045,3485,7616,0886,3116.1 (5.2–6.9)
  Women8,6428,9619,0899,6509,6373.0 (2.3–3.6)
 ≥75      
  Men2,7543,0093,4783,5773,7358.0 (6.8–9.1)
  Women7,3098,1899,0479,1169,6176.6 (5.9–7.3)
Subtotal      
 Men12,32913,15214,31215,48715,9447.0 (6.4–7.5)
 Women21,67723,17124,20925,42825,8724.5 (4.1–4.9)
 Incidence per 100,000      
  Men61.265.270.174.775.95.8 (5.2–6.3)
  Women97.0103.7107.3111.3112.03.6 (3.2–4.0)
Total34,00636,32338,52140,91541,816+5.4 (5.1–5.8)
 Incidence per 100,00080.085.489.694.094.8+4.4 (4.1–4.7)
Table 2. Number of knee arthroplasties performed in Italy between 2001 and 2005 and APC shown by sex and age groups*
 20012002200320042005APC (95% CI)
  • *

    APC = annual percentage change; 95% CI = 95% confidence interval.

Age group, years      
 25–44      
  Men859515317116018.8 (12.5–25.5)
  Women748511212412013.7 (6.9–20.9)
 45–64      
  Men1,2071,4891,7382,2512,42619.5 (17.7–21.3)
  Women3,1473,6394,4115,3885,65616.6 (15.5–17.7)
 65–74      
  Men3,3793,7934,5715,2045,80014.9 (13.8–15.9)
  Women10,56011,93113,20315,37315,75511.0 (10.4–11.6)
 ≥75      
  Men1,7822,1732,6232,8883,48017.4 (16.0–18.9)
  Women6,5177,7369,0999,67410,65412.5 (11.7–13.2)
Subtotal      
 Men6,4537,5509,08510,51411,86616.6 (15.8–17.3)
 Women20,29823,39126,82530,55932,18512.4 (12.0–12.8)
 Incidence per 100,000      
  Men32.037.444.550.756.515.2 (14.5–16.0)
  Women90.8104.6118.9133.8139.311.4 (11.0–11.8)
Total26,75130,94135,91041,07344,051+13.4 (13.1–13.8)
 Incidence per 100,00062.972.783.594.399.9+12.3 (11.9–12.7)

Women represented the majority of patients undergoing TJA procedures (average female:male ratio 1.7:1 for hip arthroplasties and 2.9:1 for knee arthroplasties). When analyzing data by age groups, most of the patients were comprised in the age groups 65–74 and ≥75 years, although the highest increases were observed in younger patients ages <65 years both for hip and knee arthroplasties (Tables 1 and 2). In 2005, we also recorded a total of 6,387 prosthesis revisions following hip arthroplasties (APC +4.9; 95% CI 4.0–5.7) and 2,295 revisions of knee arthroplasties (APC +17.4; 95% CI 15.7–19.2), as shown in Tables 3 and 4.

Table 3. Number of hip prosthesis revisions performed in Italy between 2001 and 2005 shown by sex and age groups (due to small numbers of cases, APC is shown for the total)*
 20012002200320042005APC (95% CI)
  • *

    APC = annual percentage change; 95% CI = 95% confidence interval.

Age group, years      
 25–44      
  Men5286709594 
  Women6973818289 
 45–64      
  Men263416450447499 
  Women682713658765756 
 65–74      
  Men517758751772799 
  Women1,4521,5391,5441,4831,616 
 ≥75      
  Men430595647689754 
  Women1,6231,7341,7381,7791,780 
Subtotal      
 Men1,2621,8551,9182,0032,146 
 Women3,8264,0594,0214,1094,241 
 Incidence per 100,000      
  Men6.39.29.49.710.2+9.8 (8.3–11.4)
  Women17.118.217.818.018.4+17.1 (15.1–19.2)
Total5,0885,9145,9396,1126,387+4.9 (4.0–5.7)
 Incidence per 100,00012.013.913.814.014.5+3.8 (3.0–4.7)
Table 4. Number of knee prosthesis revisions performed in Italy between 2001 and 2005 shown by sex and age groups (due to small numbers of cases, APC is shown for the total)*
 20012002200320042005APC (95% CI)
  • *

    APC = annual percentage change; 95% CI = 95% confidence interval.

Age group, years      
 25–44      
  Men1426273126 
  Women161718817 
 45–64      
  Men6013110282131 
  Women126271188281271 
 65–74      
  Men131260221228260 
  Women417838593777838 
 ≥75      
  Men790132155142 
  Women3030434453610 
Subtotal      
 Men284417482496559 
 Women8621,1261,2331,5191,736 
 Incidence per 100,000      
  Men1.42.12.42.42.7+13.9 (10.6–17.4)
  Women3.95.05.56.67.5+1.3 (0.3–2.3)
Total1,1461,5431,7152,0152,295+17.4 (15.7–19.2)
 Incidence per 100,0002.73.64.04.65.2+16.2 (14.5–18.0)

The incidence per 100,000 inhabitants increased from 80 to 94.8 for hip arthroplasties (APC +4.4; 95% CI 4.1–4.7) and from 62.9 to 99.9 for knee arthroplasties (APC +12.3; 95% CI 11.9–12.7), with the highest increases being observed in men for both TJAs (APC +5.8; 95% CI 5.2–6.3 and APC +15.2; 95% CI 14.5–16.0, respectively). An increase in the incidence per 100,000 inhabitants was also shown for TJA revisions, from 12.0 to 14.0 per 100,000 (APC +3.8; 95% CI 3.0–4.7) and from 2.7 to 4.6 per 100,000 (APC +16.2; 95% CI 14.5–18.0) for hip and knee arthroplasties, respectively. Notably, the incidence per 100,000 hip prosthesis revisions showed higher increases in women (APC +17.1; 95% CI 15.1–19.2).

As shown in Tables 1 and 2, approximately 30% of hip arthroplasties and 20% of knee arthroplasties were performed annually in patients ages <65 years, thus affecting the working population. The mean LOS for hip arthroplasty without complications progressively decreased, from 14.6 days in 2001 to 12 days in 2005 (13.6 days in 2002, 13 days in 2003, and 12.5 days in 2004). The mean LOS for knee arthroplasty without complications decreased, from 14 days in 2001 to 11 days in 2005 (13.5 days in 2002, 12.1 days in 2003, and 11.5 days in 2004). The LOS observed for private hospitals is lower than that reported for public hospitals, although the latter hospitalizes the majority of patients. The average LOS was 15 days for hip prosthesis revision and 14.5 days for knee prosthesis revision. The mean LOS was 10 days in the case of postoperative infection, 15 days for deep vein thrombosis, 14 days for pulmonary embolism, and 20 days for in-hospital rehabilitation. Rehabilitation carried out at a patient's home was computed in 30 days, although it can be extended up to 60 or 90 days in some cases. As shown in Table 5, the total loss of working days following hip and knee arthroplasty has been estimated as approximately 805,000 days (2001) and 1 million days (2005). Most working days were lost because of hip arthroplasty. The computation of working days lost was limited to patients ages <65 years and included hospitalization, complications, and rehabilitation. Despite a notable reduction in the average hospital LOS for both hip and knee arthroplasties across the 5 examined years, the number of working days lost increased from 2001 to 2005, as a consequence of the increase in the number of TJA procedures.

Table 5. Estimated working days lost by patients ages <65 years and related APC*
Total joint arthroplasties20012002200320042005APC (95% CI)
  • *

    APC = annual percentage change; 95% CI = 95% confidence interval.

Hip arthroplasties545,437566,146570,319631,409631,351+4.1 (4.0–4.2)
Knee arthroplasties259,910262,801300,761367,413386,586+12.0 (11.9–12.2)
Total days lost805,347828,947871,080998,8221,017,937+6.8 (6.7–6.9)

Hospital costs sustained by the national health care system for TJA procedures increased from 741 million euros to 1 billion euros between 2001 and 2005 (Table 6). Among those costs, hip arthroplasty hospital costs increased from 412 to 538 million euros and knee arthroplasty costs increased from 329 to 517 million euros over the 5-year period. Rehabilitation costs were assessed as 228 and 322 million euros in 2001 and 2005, respectively (Table 6). Overall estimated costs of postoperative complications (including deep vein thrombosis, pulmonary embolism, and infections) were computed between 3.1 (2001) and 4.4 million euros (2005), assuming a 0.5% incidence rate per each type of complication (Table 6). Based on 2005 data, the average cost per patient (including hospital, rehabilitation, and complication costs) was assessed as 16,835 and 15,358 euros for hip and knee arthroplasty, respectively. Considering the overall 8-year study period, total costs increased up to 6 billion euros, going from 973 million euros in 2001 to 1.4 billion euros in 2008. Mortality following TJA resulted in 335 deaths in 2001, 373 deaths in 2002, 410 deaths in 2003, 450 deaths in 2004, and 473 deaths in 2005, thus representing a cause of death comparable with other diseases characterized by low mortality such as tuberculosis (almost 400 deaths per year in Italy) (9).

Table 6. Estimated costs (million euros) of hip/knee arthroplasties in Italy between 2001 and 2005
 20012002200320042005Total 2001–2005
Hospital costs      
 Hip4124894815005382,420
 Knee3293784105105172,144
 Total7418678911,0101,0554,564
Rehabilitation      
 In hospital      
  Hip4043454849225
  Knee2833384447190
 Home based      
  Hip93101106113115528
  Knee677890103111449
 Total2282552793083221,392
Deep vein thrombosis      
 Hip0.50.50.50.60.62.7
 Knee0.30.40.50.50.62.3
 Total0.80.911.11.25
Pulmonary embolism      
 Hip0.80.90.9113.6
 Knee0.60.70.80.90.93.9
 Total1.41.61.71.91.98.5
Postoperative infections      
 Hip0.50.50.60.60.72.9
 Knee0.40.40.50.60.62.1
 Total0.90.91.11.21.35.4
Overall costs      
 Hip5476356346637043,182
 Knee4254905406596772,791
 Total9721,1261,1741,3221,3815,973

DISCUSSION

This is the first Italian study addressing the issue of TJAs performed because of symptomatic OA. We considered these surgical procedures to closely reflect the incidence of severe OA, although a proportion of affected patients do not undergo surgery. Other limitations of this study are mostly attributable to the estimation rates concerning prosthesis revision postoperative infections, deep vein thrombosis, pulmonary embolism, and mortality (based on the Tuscany regional database and international literature). However, these complications and mortality rates after TJA are known to be rare in developed countries, where high hospital standards are adopted (13–35). Therefore, our analyses are thought to be only minimally influenced by a possible variability of complications and mortality rates. A prospective observational study of unselected operations carried out in the UK by Williams et al in 2002 reported a 3-month mortality rate between 0.4% and 0.7% (28). Zhan and colleagues reported that approximately 200,000 total hip replacements, 100,000 partial hip replacements, and 36,000 prosthesis revisions were performed in the US in 2003, with intrahospital mortality rates of 0.33%, 3.04%, and 0.84%, respectively (29). In 2006, Doro et al reported mortality following these surgical procedures to be 0.16–0.29% (30). Mahomed et al showed that the rates of complications occurring within 90 days after primary total hip replacement were likely to be 1.0% for mortality, 0.9% for pulmonary embolism, and 0.2% for wound infection (19). Blom and colleagues assessed early postoperative mortality in an unselected consecutive series of 1,727 primary THAs where patients had not routinely received chemothromboprophylaxis, reporting a 3-month mortality rate of 1% (31). More recently, Liu et al have analyzed trends in characteristics of THA in the US through the National Hospital Discharge Survey from 1990 to 2004 for trends concerning in-hospital mortality and hospital LOS (32). According to this study, the number of THAs performed increased by 158%, while mortality rates remained low and decreased slightly (from 0.32% to 0.29%) (32). The prevalence of procedure-related complications decreased over time and the LOS decreased from an average of 8.7 to 4.5 days (32). Similarly, Aynardi et al assessed the 90-day mortality rate after THA by retrospectively reviewing 7,478 consecutive patients undergoing primary or revision THA between January 2000 and July 2006, reporting overall 30- and 90-day mortality rates of 0.24% and 0.55%, respectively (33). In Italy, Sperati and colleagues analyzed regional hospital discharge database and mortality data in public and private Italian hospitals in Lazio, reporting a global mortality rate after THA of 1.4% (30-day mortality 1.3% and 90-day mortality 2.2%) (34).

Our results showed a remarkable increase in the number of TJAs performed from 2001 through 2005 because of OA. This finding seems to be consistent with a long-term US survey that observed a progressive rise in the number of arthroplasties between 1990 and 2002 (35). According to these data, the rate of primary total arthroplasties per 100,000 inhabitants in the US increased by approximately 50% over 13 years (35). The number of procedures increased much more in people ages 45–64 years. The analysis of the US Nationwide Inpatient Survey from 1997 to 2004 revealed that approximately 225,900 hip arthroplasties were performed during 2004, corresponding to a 37% increase compared with the year 2000, but it is estimated that nearly 600,000 hip arthroplasties will be performed in the US by the year 2015 (36). Another Spanish study has documented the increase in the rate of hip arthroplasties per 10,000 inhabitants, which increased from 4.1 to 6.6 in Barcelona and Catalonia between 1994 and 2000 (37).

We believe that the analysis of the cost of DRGs for TJAs, complications, and rehabilitation can be considered as a reliable proxy of general costs sustained by the national health care system. We have estimated that total costs sustained for TJAs increased from 973 million to 1.4 billion euros, with the average cost per patient being 16,835 and 15,358 euros for hip and knee arthroplasties, respectively. Stargardt assessed the costs of primary hip arthroplasties performed in 42 hospitals of 9 European Union countries, finding that the total cost for THAs ranged from 1,290 euros (Hungary) to 8,739 euros (The Netherlands), with a mean cost of 5,043 euros (38). In that study, Italy was situated among the countries with the highest costs, as confirmed by our results. In a similar French study, it was found that 73,150 hospital admissions in the year 2001 resulted in total hospital costs of 591 million euros, thus being comparable with our Italian results (39). The loss of working days by people ages <65 years has been estimated at approximately 1 million days in 2008. Although surgery is the gold standard treatment in the case of severe symptomatic hip OA, other treatment options may also be considered in less severe cases, when medical therapy or conservative strategies allow an acceptable quality of life for the patients. The World Health Organization recommendations on OA (40) focus on primary prevention and foster the adoption of all possible conservative treatments before undergoing surgery, unless quality of life and function impairment become unacceptable. The availability of effective conservative treatment (such as eco-guided infiltration of the hip and knee joints) could prolong the survival of the patient's joint, thus possibly reducing loss of productivity, NSAID expenditures, and also the need for revision surgery (given that the life of a prosthesis does not currently exceed 20 years). The number of revisions is progressively increasing after both hip and knee arthroplasties.

Concerning the prosthesis revision rate, the study performed by Williams et al in 2002 showed that 2.6% of patients underwent another operation on the same hip within 1 year (28). Dixon et al reported, between 1991 and 2000, an increasing incidence of primary THA (+18%), while the incidence of THA revisions was almost double. Over the 10 years, the proportion of THA procedures requiring revision rose from 8% to 20% (41). Moreover, Lübbeke and colleagues showed that unadjusted quality of life and satisfaction were significantly lower after revision (42). Adjustment for patient characteristics revealed that this difference was only partially explained by the greater morbidity and older age of the patients undergoing revision (42). A study conducted by Ong et al reported that the average economic burden of prosthesis revision reached 18.8% (range 17.4–20.2%) of THAs (43). Recent data from the New Zealand Joint Registry of THAs performed between 1999 and December 2006 reported 920 revisions (2.16%) of 42,665 primary THAs (44). Bozic et al evaluated the mechanisms of surgical failure and the types of revision generally needed after THA procedures performed in the US (45). This analysis was carried out by using the US ICD-9-CM databases concerning codes specifically related to revision after THA in a large sample representative of the US population until 2006. The study found that the most common type of revision after a THA procedure was the all-component revision (41.1%), while the most common causes of revision were instability/dislocation (22.5%), mechanical loosening (19.7%), and infection (14.8%) (45). Revision procedures following THA were most commonly performed in large, urban, nonteaching hospitals for Medicare patients in people ages 75–84 years. The average hospital LOS for all types of revision arthroplasties was 6.2 days, and the average total charges were found to be $54,553 (46).

Our study confirms that the socioeconomic burden of TJAs (including revision surgery) due to hip and knee OA is growing and heavily affecting the working population.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Piscitelli had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Piscitelli, Iolascon, Di Tanna, Bizzi, Chitano, Argentiero, Neglia, Giolli, Distante, Gimigliano, Brandi, Migliore.

Acquisition of data. Piscitelli, Iolascon, Di Tanna, Bizzi, Chitano, Argentiero, Neglia, Giolli, Distante, Gimigliano, Brandi, Migliore.

Analysis and interpretation of data. Piscitelli, Iolascon, Di Tanna, Bizzi, Chitano, Argentiero, Neglia, Giolli, Distante, Gimigliano, Brandi, Migliore.

Ancillary