United States Hip and Knee Replacement Cohort
The United States total hip arthroplasty and total knee arthroplasty
cohorts were identified from the 5% systematic sample of Medicare
beneficiaries from January 1, 1997 to December 31, 2004. Elective primary and
revision hip or knee surgery were identified with ICD-9-CM and Current
Procedural Terminology-4 (CPT-4) codes. The unique Medicare beneficiary
identification number was used to follow patients longitudinally between
primary and revision surgery. Survival of the implant was from the date of the
primary surgery until the date of revision. The following exclusion criteria
were applied: (1) patients who were younger than sixty-five years, (2) a total
hip arthroplasty or total knee arthroplasty associated with traumatic
fracture, (3) patients who were not enrolled in both Parts A and B of
Medicare, and (4) enrollees in health maintenance organizations.
Norwegian Total Hip Arthroplasty and Total Knee Arthroplasty and
Swedish Total Hip Arthroplasty Cohorts
The Norwegian total hip arthroplasty and total knee arthroplasty cohorts
and the Swedish total hip arthroplasty cohort were extracted from their
respective national arthroplasty registries, and all patients who were at
least sixty-five years old and had a primary total hip or total knee
arthroplasty during the same time-period were included. Total knee
arthroplasty data from the Swedish registry were unavailable for analysis.
Patients who did not receive a revision before the end of the follow-up
period, died without receiving a revision, or were lost to follow-up (e.g.,
emigrated) were considered censored, and the longevity of their implants was
calculated to the date of death, date of emigration, or the end of the study,
whichever was earliest.
Statistical Analysis
The overall survival probabilities for the United States, Norwegian, and
Swedish cohorts were assessed with use of the Kaplan-Meier method, and Cox
regression was used to evaluate the effects of patient attributes on survival
(a p value of <0.05 indicated significance).
Economic Impact
Medicare claims data (1997 through 2004) were used to identify primary and
revision arthroplasties. ICD-9-CM and CPT-4 codes were used to extract the
records from Part-A (hospital) and Part-B (physician and/or surgeon) claims,
respectively. The historical prevalence of primary and revision total hip
arthroplasty and/or total knee arthroplasty was calculated for the subgroups
of patients in the United States and was independently modeled to vary over
time with use of Poisson
regression2. The
overall projected economic impact of revision total hip arthroplasty and total
knee arthroplasty through 2015 for hospital and surgeon charges was determined
by combining the projected adoption of the procedures, changes in procedural
charges, and changes in the patient population, and it accounted for
inflation.
Infections
Historically, the number of infections after revision total knee
arthroplasty has been greater than the number of infections after revision
total hip arthroplasty (Fig.
1). The number of infections after revision total hip arthroplasty
was projected to increase from 3400 in 2005 to 46,000 in 2030, while
infections after revision total knee arthroplasty were projected to increase
from 6400 in 2005 to 175,500 in 2030. The incidence of deep infection as a
reason for revision was projected to increase at a faster rate for total knee
arthroplasty than for total hip arthroplasty
(Fig. 2). Revisions of total
hip arthroplasties done because of deep infection were projected to increase
from 8.4% in 2005 to 47.5% in 2030. Similarly, revisions of total knee
arthroplasties because of deep infection were projected to increase from 16.8%
in 2005 to 65.5% in 2030. The economic burden of infections is expected to
exceed 50% of the inpatient resources available for revisions by 2025 for
total hip arthroplasty and by 2016 for total knee arthroplasty. The incidence
of deep infection was projected to exceed 50% after 2030 for total hip
arthroplasty and by 2022 for total knee arthroplasty. The overall incidence of
deep infection (the infection burden) was projected to increase at a
comparable rate for both primary and revision total hip arthroplasty and total
knee arthroplasty (Fig. 3),
with the infection burden increasing between 2005 and 2030 from 1.4% to 6.5%
for total hip arthroplasty and from 1.4% to 6.8% for total knee
arthroplasty.
Survival
By the end of 2004, based on 5% Medicare sample, 3.6% (979) of 27,076
primary hip arthroplasties done in the United States required revision
(Table I). In contrast, 1.8%
(1451) of 82,037 primary hip arthroplasties in Sweden and 2.2% (937) of 41,823
primary hip arthroplasties in Norway were revised. By the end of 2004, 2.1%
(1154) of 55,519 primary knee implants in the United States compared with 3.8%
(532) of 13,969 primary knee implants in Norway required revision. By the end
of the study period, 93.0% of the total hip arthroplasties and 95.9% of the
total knee arthroplasties done in elderly Medicare patients in the United
States remained free of revision. The revision rate for total hip arthroplasty
was found to be significantly higher than that for total knee arthroplasty
(adjusted hazard ratio = 1.77, p < 0.0001) (Figs.
4 and
5). In contrast, the rate of
revision-free survival of total hip arthroplasties was 96.8% in Sweden and
96.4% in Norway, which was significantly higher than that in the United States
(p < 0.0001). The rate of revision-free survival of total knee
arthroplasties was 93.9% in Norway, which was significantly lower than that in
the United States (p < 0.001).
Age was consistently found to be a highly significant factor (p <
0.004), as Medicare patients had the primary surgery at a younger age and
needed substantially more revisions (Fig.
6). For the United States patients, a significant difference was
found between men and women with regard to the survival of knee implants (p =
0.02) but not with regard to the survival of hip implants (p = 0.17)
(Fig. 7). The percentage of
revisions with matching laterality (matched left-right side between the
primary and revision surgeries) was only 42.2% for hips and 47.9% for knees.
The rate of hip implants that remained free of revision was comparable between
those with known or unknown laterality (p = 0.11). This was also similar for
knee implants (p = 0.84).
Economic Impact
Annual hospital charges were estimated to increase between 2005 and 2015 by
340% to $17.4 billion for primary total hip arthroplasty and by 450% to $40.8
billion for primary total knee arthroplasty
(Fig. 8). Corresponding
surgical charges were estimated as $1.9 billion (a 180% increase) for total
hip arthroplasty and $5.06 billion (a 250% increase) for total knee
arthroplasty in 2015. Hospital charges for revision total hip arthroplasty and
revision total knee arthroplasty were projected to increase by 290% to $3.8
billion and by 450% to $4.1 billion (Fig.
9). Corresponding surgical charges were estimated as $0.36 billion
(a 140% increase) for revision total hip arthroplasty and $0.34 billion (a
160% increase) for revision total knee arthroplasty in 2015. Approximately
two-thirds of the charges for these procedures were allocated to women, except
for revision total knee arthroplasty (Table
II). The proportion of charges for the age-group of sixty-five to
sixty-nine years old increased substantially, with corresponding decreases in
the remaining age-groups (Table
III).