Introduction

Acetabular fractures in older patients are occurring more frequently and are associated with high rates of mortality and non-fatal complications [1,2,3]. These older patients often have pre-existing hip osteoarthritis, osteoporosis, and fracture characteristics predictive of post-traumatic arthritis [4, 5]. Surgical treatment options for older patients with acetabular fracture include open reduction internal fixation (ORIF), total hip arthroplasty (THA), or the combined used of ORIF and THA (ORIF + THA) [6, 7].

While ORIF has been the historical standard of care [6, 8, 9], recent evidence suggests that acute THA may provide better functional outcomes with fewer reoperations in older patients with acetabular fracture [10,11,12,13,14,15,16,17]. It is not known whether surgeons who treat acetabular fractures have changed their practice to perform arthroplasty more frequently. Our purpose is to investigate if the acute use of THA for acetabular fractures in middle-aged and older adults—patients for whom an elective total hip arthroplasty would represent a reasonable intervention for osteoarthritis—has increased over time. We hypothesize that the relative incidence of THA and ORIF + THA have increased in the last decade compared with ORIF as a standard of care. We further characterize facility, patient, and fracture characteristics that may be associated with the choice to manage acetabular fracture with THA as a primary treatment or with adjunctive ORIF.

Materials and methods

Data source

Adults ≥ 45 years of age treated for acetabular fracture within 3 weeks of admission between 2010 through 2020 were retrospectively identified from the Nationwide Inpatient Sample (NIS). The NIS is curated by the Healthcare Cost and Utilization Project (HCUP) and includes data on over 7 million hospital discharges per year [18, 19]. In 2012 the NIS included a 20% stratified sample of inpatient admissions across the United States, while prior to 2012 the NIS sampled 20% of discharges from participating hospitals. This methodologic change was intended to improve the precision of national estimates by reducing sampling error [20]. Deidentified publicly available data were used for this investigation, thus this study was exempt from Institutional Review Board review.

Study population

International Classification of Diseases Ninth Edition Clinical Modification (ICD-9-CM) and Tenth Edition Clinical Modification (ICD-10-CM) codes were used to identify patients with acetabular fractures (Supplementary Table S1). Operative treatments including open reduction internal fixation (ORIF), total hip arthroplasty (THA), and combined ORIF and THA (ORIF + THA) were identified using ICD-9 and ICD-10 procedure codes (PCS; Supplementary Table S2). Patients < 45 years old and those without recorded operative procedure or procedures > 3 weeks after admission were excluded to capture patients with acute injuries and those who are most likely to benefit from and be treated by acute THA [21,22,23].

A weighted sample estimate representing 22,795 adults met inclusion criteria (Supplementary Fig. S1). The relative incidence of each treatment was 80.9% ORIF, 12.1% THA, and 7.0% ORIF + THA. Patients receiving each treatment significantly differed by age, sex, insurance type, hospital size, hospital type, and comorbidities overall (Tables 1, 2, Supplemental Table S3-S4).

Table 1 Patient characteristics, length of stay, and costs of acetabular fracture surgery by procedure (ages ≥ 65 years)
Full size table
Table 2 Patient characteristics, length of stay, and costs of acetabular fracture surgery by procedure (ages 45–64 years)
Full size table

Study outcomes

The primary outcome was annual change in the incidence of ORIF, THA, and ORIF + THA. Secondary outcomes included hospital length of stay and national charge estimates for cost of hospital stay.

Statistical methods

Annual procedure incidence was modeled using linear regression on proportional changes in yearly cases. Patients were stratified by age 45–64 years (middle-aged) or ≥ 65 years (older adult). Further subgroup analysis considered the effects of sex and insurance type within age groups on treatment. Chow tests were conducted to determine differences in linear regression coefficients β across groups. Adjusted R2 was used to determine goodness-of-fit. Continuous and categorical variables were compared across unweighted treatment groups using one way analysis of variance and chi-squared, respectively. National charge estimates for hospital stay were derived using the NIS discharge weighting scheme which reflects a ratio of national discharges to discharges captured by NIS. Statistical analyses were performed using Stata Version 17.0 (College, Station, TX), reporting 2-sided p values with the level of significance for p < 0.050.

Results

Treatment by age

For patients aged ≥ 65 years, 5860 (65.8%) underwent ORIF, 1844 (20.8%) underwent THA, and 1162 (13.1%) underwent ORIF + THA. For patients aged 45–64 years, 12,588 (90.4%) underwent ORIF, 906 (6.5%) underwent THA, and 435 (3.1%) underwent ORIF + THA. We observed greater overall use of THA (20.8%) and ORIF + THA (13.1%) for patients ≥ 65 years with acetabular fracture versus greater use of ORIF (90.4%) in the 45–64 group (Table 1). The mean age of the ≥ 65 group was 76.1 years (95% confidence interval (CI) 75.8–76.5 years), with those undergoing ORIF (75.5 years; 95% CI 75.0–75.9 years) being significantly younger than those undergoing THA (77.6 years; 95% CI 77.6–78.3 years) or ORIF + THA (77.2 years; 95% CI 76.3–78.0 years) (p < 0.001) (Table 1). The mean age of the 45–64 group was 54.9 years (95% CI 54.7–55.1 years), with significantly younger patients undergoing ORIF (54.6 years; 54.3–54.8 years) than those undergoing THA (57.6 years; 95% CI 56.7–58.4 years) or ORIF + THA (58.9 years; 95% CI 57.8–60.0 years) (Table 2).

Treatment trends by age

We observed an increase in the use of THA for acetabular fracture over the study period in middle-aged and older adults. From 2010 to 2020, the use of ORIF for patients ≥ 65 years decreased 21.5% [R2 = 0.75; β1 = −2.4% per year (95% CI −3.4 to −1.3%)], while THA increased 16.5% [R2 = 0.87; β1 = 1.7% per year (95% CI 1.2–2.2%)] and ORIF + THA increased 5.0% [R2 = 0.38, β1 = 0.6% per year (95% CI 0.0–1.3%)] (Fig. 1a). Among patients aged 45–64 years, the use of ORIF decreased 7.4% [R2 = 0.75; β1 = −0.9% per year (95% CI −1.2 to −0.5%)], while THA increased 4.8% [R2 = 0.62; β1 = 0.6% per year (95% CI 0.2–0.9%)] and ORIF + THA increased 2.6% [R2 = 0.73; β1 = 0.3% per year (95% CI 0.2–0.4%)], (Fig. 1b). Annual rates of change in practice were significantly different between age groups for ORIF (p < 0.001) and THA (p < 0.001), but not for ORIF + THA (p > 0.05).

Fig. 1
figure 1

Acetabular treatment trends by open reduction internal fixation (black), total hip arthroplasty (dark grey), and open reduction internal fixation + total hip arthroplasty (light grey) in the NIS database from 2010 to 2020. A In ages ≥ 65.  B In ages 45–64

Full size image

Treatment trends by sex

Patient sex was associated with changes in practice over time. Acute THA for acetabular fracture significantly increased over time among women ≥ 65 years [R2 = 0.68; β1 = 2.5% per year (95% CI 1.2–3.8%)]. Treatment by ORIF and ORIF + THA did not change over time among women ≥ 65 years. No temporal change of practice was observed among men [R2 = 0.24; β1 = 0.5% per year (95% CI −0.2 to 1.2%); p = 0.002; Fig. 2a, b]. No differences in treatment by sex were observed over time among patients aged 45–64 years (p > 0.05; Fig. 2c, d).

Fig. 2
figure 2figure 2

Acetabular treatment trends by open reduction internal fixation (black), total hip arthroplasty (dark grey), and open reduction internal fixation + total hip arthroplasty (light grey) in the NIS database from 2010 to 2020. A males in ages ≥ 65. B Females in ages ≥ 65.  C Males in ages 45–64. d Females in ages 45–64

Full size image

Treatment trends by insurance type

Medicare insurance status was associated with the greatest growth in use of THA in acetabular fracture among patients aged ≥ 65 years [R2 = 0.80; β1 = 2.0% per year (95% CI = 1.2–2.7%) vs R2 = 0.14; β1 = 0.6% per year (95% CI = −0.5 to 1.6%); p = 0.006]. Additionally, the relative use of ORIF declined more rapidly over time in the Medicare insured patients (R2 = 0.67; β1 = −3.0% per year (95% CI = −4.7 to −1.4%)) when compared to non-Medicare insured patients [R2 = 0.15; β1 = −0.6% per year (95% CI = −1.7–0.5%); p = 0.03]. No differences in longitudinal ORIF + THA use rates were observed between insurance groups (p > 0.05; Fig. 3a, b). Insurance type was not significantly associated with treatment in patients aged 45–64 years (p > 0.05; Fig. 3c, d).

Fig. 3
figure 3figure 3

Acetabular treatment trends by open reduction internal fixation (black), total hip arthroplasty (dark grey), and open reduction internal fixation + total hip arthroplasty (light grey) in the NIS database from 2010 to 2020. A Medicare insured in ages ≥ 65. B Non-Medicare insured in ages ≥ 65. C Privately insured in ages 45–64. d Non-privately insured in ages 45–64

Full size image

Treatment trends by hospital type

Hospital type was associated with treatment but not significantly associated with changes in practice over time. The incidence of THA and THA + ORIF increased over time in rural, urban non-teaching, and urban teaching hospitals over time, but the rates of practice change were not significantly different across hospital types among patients aged ≥ 65 years (Supplementary Fig. S2a–S3c) or patients 45–64 years old (Supplementary Fig. S2d–S3f).

Trends in length of stay by treatment

Length of stay was significantly shorter among patients who received THA [6.4 days (95% CI 5.8–7.0 days)] versus patients who received ORIF [9.7 days (95% CI 9.4–9.9 days); p < 0.001) or ORIF + THA (8.2 days (95% CI 7.3–9.0 days); p = 0.004]. Length of stay did not significantly vary over time by procedure among patients ≥ 65 years old with acetabular fracture (Fig. 4a). In patients aged 45–64 years, the overall hospital length of stay decreased over time for ORIF [R2 = 0.65; β1 = −0.17 days per year (95% CI −0.27 to −0.01 days)] but remained unchanged for both THA [R2 = 0.05; β1 = −0.11 days per year (95% CI = −0.5, to 0.27 days)] and ORIF + THA [R2 = 0.24; β1 = −0.22 days per year (95% CI −0.5 to 0.1 days)] (Fig. 4b).

Fig. 4
figure 4

Total hospital length of stay trends in days by open reduction internal fixation (black), total hip arthroplasty (dark grey), and open reduction internal fixation + total hip arthroplasty (light grey) in the NIS database from 2010 to 2020. a In ages ≥ 65. b In ages 45–64

Full size image

Trends in estimated charges by treatment

National estimated hospital charges were significantly lower in patients who received THA [$112,591.40 (95% CI $104,912.40–$120,270.40)] compared to patients who received ORIF [$143,542.80 (95% CI $138,922.50–$148,163.10)] (p < 0.001) and ORIF + THA [$159,730.10 (95% CI $144,901.70–$174,558.60)] (p < 0.001). In patients ≥ 65 years old, the cost of both ORIF [R2 = 0.47; β1 = $2285.59 per year (95% CI $456.96–$4114.22)] and ORIF + THA [R2 = 0.37; β1 = $5308.89 per year (95% CI $75.24–$10,542.53)] increased significantly per year, while the cost of THA [R2 = 0.01; β1 = $550.01 per year (95% CI −$3844.19–$4944.22)] did not change (Fig. 5a). In patients 45–64 years old, no differences in cost of ORIF [R2 = 0.19; β1 = $1762.39 per year (95% CI −$1062.08–$4586.87)], THA [R2 = 0.13; β1 = $2,272.71 per year (95% CI −$2234.15–$6779.56)], or ORIF + THA [R2 = 0.00; β1 = $245.90 per year (95% CI −$7443.65–$7935.46)] were observed over time (Fig. 5b).

Fig. 5
figure 5

Inflation adjusted total patient cost trends in dollars by open reduction internal fixation (black), total hip arthroplasty (dark grey), and open reduction internal fixation + total hip arthroplasty (light grey) in the NIS database from 2010 to 2020. a In ages ≥ 65. b In ages 45–64

Full size image

Discussion

We observed substantial practice changes in the surgical management of middle-aged and older US adults with acetabular fracture between 2010 and 2020 in the Nationwide Inpatient Sample. Among older adults aged ≥ 65 years, the incidence of THA and ORIF + THA increased 16.5% and 5.0%, respectively, while the use of ORIF decreased by 21.5%. Practice changes were less substantial among middle-aged adults aged 45–64 years: the incidence of THA and ORIF + THA increased 4.8% and 2.6%, respectively, while the use of ORIF decreased 7.4%. The practice of acute THA for acetabular fracture grew 1.6% annually among older adults and 0.6% annually among middle-aged adults. To our knowledge, these novel observations reflect progressive adoption of arthroplasty for acetabular fractures in middle-aged and older adults.

Acetabular fractures in older adults incur substantial morbidity, disability, and healthcare costs [2, 3]. The adoption of acute THA as either a primary treatment or combined with adjunctive ORIF may reflect a response by surgeons to contemporary evidence for benefits to arthroplasty over ORIF alone in older patients. In a recent systematic review on the outcomes following acetabular fracture in 3928 patients aged 55–99 years, McCormick et al. found ORIF was associated with 2.2 greater odds of non-fatal complications versus THA and a pooled 15% rate of conversion to THA after ORIF at a mean of 29.4 months [12]. Although early reports of ORIF + THA have reported concerning rates of heterotopic ossification, dislocation, and other mechanical failures, single center series suggest improved survival with low rates of reoperation [7, 24]. Manson et al. prospectively randomized 47 patients over age 60 to either ORIF or ORIF + THA and reported a 28% absolute risk reduction in overall reoperations and 12% greater Harris Hip score with no dislocations in either group within 2 years [22]. Borg et al. found improved three-year hip survival in patients undergoing ORIF + THA when compared to ORIF alone [13]. Lin et al. noted equivalent functional outcomes to patients undergoing primary THA for osteoarthritis with a 15% complication rate [17]. However, Kelly et al. identified a 27% surgical complication rate, 13% medical complication rate, and 18% all cause revision risk in a recent review of insurance claims data [25]. These studies provide compelling, if tempered, support for the use of arthroplasty to treat acetabular fractures with a reasonable probability of a favorable outcome in older patients [24].

ORIF remains widely regarded as the standard of care for younger patients with displaced or minimally-displaced but unstable acetabular fractures. We did not observe a clinically relevant change—a minimal 0.9% decline—in the incidence of ORIF for acetabular fracture in middle-aged patients aged 45–64 years. Younger and middle-aged adults experience excellent hip survival and functional outcomes when the congruity and stability of the hip joint are restored [5, 6]. However, specific fracture patterns and patient factors may increase the risk of subsequent posttraumatic arthritis requiring conversion to THA [8, 26, 27]. It is possible that specific fracture and patient characteristics, evolution in implant materials and bearing surfaces, and surgeon confidence in implant survivorship may have driven the small increases in the use of THA (4.8%) and ORIF + THA (2.6%) in middle-aged adults.

We observed differences in the treatment of acetabular fractures by patient sex among older adults: women over age 65 were more likely to receive a THA for acetabular fracture than men, with the incidence of THA for acetabular fracture increasing 2.5% annual in women and 0.5% per year in men. Disparities in treatment by sex may be related to sex-based differences in underlying risk factors for poor outcome with ORIF – chiefly, perhaps, osteoporosis. Most acetabular fractures in the elderly are associated with osteopenic or osteoporotic bone and consequently demonstrate more complex fracture patterns with impaction, comminution, and higher rates of failure when treated with ORIF [5, 26]. Women are more likely to develop osteoporosis than men with advancing age, but the burden of metabolic bone disease among older men and the implications for the outcomes of acetabular ORIF should not be disregarded [28, 29]. The sex-based differences in treatment we observed may also reflect survivorship bias as more women survive to old age and sustain acetabular fractures than men. We did not observe longitudinal trends in treatment by race.

Medicare primary insurance was also associated with a more rapid adoption of THA and concurrent decline in use of ORIF compared to patients without non-Medicare primary insurance. Medicare reimbursement for hip fracture declined 30% between 2000 and 2020 [30]. Medicare reimbursement of acetabular fracture ORIF under DRG 536 is also equivalent to pelvic fracture closed reduction internal fixation, despite major differences in perioperative resources, risk, blood loss, surgical time, and technical difficulty of open acetabular fracture surgery versus minimally invasive percutaneous pelvic ring screw fixation [31]. Unlike femoral neck and head fractures, THA for acetabular fracture is not eligible for bundled payment programs which may protect reimbursement for this use of THA [32]. Our data also demonstrate longitudinally stable costs of care for patients undergoing THA, while ORIF and ORIF + THA related costs rose significantly over the same period. These differential costs are important considerations and are not fully explained by inflation [33]. The combined economic incentives favoring THA may partially account for the practice change observed in patients with acetabular fracture and Medicare primary insurance.

This study has several limitations. The retrospective review of a hospital discharge database is subject to documentation and classification bias. The risk of misclassification may be potentiated by the change from ICD-9 to ICD-10 coding schemes in 2015. We attempted to address this through the careful exclusion of concomitant injuries which may have confounded our analysis. Furthermore, while the NIS database is a validated method to determine national estimates on procedural trends, it is limited to a 20% sample of the actual annual volume of inpatient care. Our observations reflect weighted national estimates rather than observations [34]. This study did not account for injury mechanism, injury severity, or comorbid conditions, which may also influence treatment. These trends may not be generalizable to practice outside of the United States.

In conclusion, THA is increasingly being used as a primary or adjunctive treatment acetabular fracture in middle aged and older US adults, while the incidence of ORIF has declined. Female sex and Medicare insurance were significantly associated with more rapid adoption of THA as a treatment for acetabular fracture in patients ≥ 65 years old. The downstream implications of these evolving treatment trends unknown and warrant further investigation, particularly with regard to physical function, hip survival, health status, independence, and costs of care.