JBJS | Thirty-Day Mortality After Elective Total Hip Arthroplasty

The Journal of Bone & Joint Surgery, Volume 83, Issue 10

Scientific Article | October 01, 2001

Thirty-Day Mortality After Elective Total Hip Arthroplasty

Javad Parvizi, MD, FRCS; Blake G. Johnson, MD; Charles Rowland, MS; Mark H. Ereth, MD; David G. Lewallen, MD

Investigation performed at the Departments of Orthopedics and Anesthesiology and the Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota
Javad Parvizi, MD, FRCS
Blake G. Johnson, MD
Charles Rowland, MS
Mark H. Ereth, MD
David G. Lewallen, MD
Departments of Orthopedics (J.P., B.G.J., and D.G.L.) and Anesthesiology (M.H.E.) and Section of Biostatistics (C.R.), Mayo Clinic and Mayo Foundation, 200 First Street S.W., Rochester, MN 55905. E-mail address for D.G. Lewallen: lewallen.david@mayo.edu

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

The Journal of Bone & Joint Surgery. 2001; 83:1524-1528

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Background: Previous reports on perioperative mortality associated with hip arthroplasty have not documented, to our knowledge, patient characteristics and surgical factors that increase the likelihood of death. The purpose of this study was to determine the prevalence of and associated risk factors for perioperative death after elective hip arthroplasty.

Methods: The records of 30,714 consecutive patients who had undergone elective hip arthroplasty at our institution from 1969 to 1997 were retrospectively reviewed to identify patients who had died within thirty days after the procedure. Mortality rates were determined according to age, gender, diagnosis, implant type, and fixation mode.

Results: Ninety deaths occurred within thirty days after elective total hip arthroplasty, for an overall mortality rate of 0.29% (ninety of 30,714). The thirty-day mortality rate was significantly higher for patients with preexisting cardiovascular disease (p < 0.0001), male patients (p < 0.0001), and patients who were seventy years of age or older (p < 0.0002). The mortality rate was slightly, but not significantly, higher for patients with an underlying diagnosis of rheumatoid arthritis (p < 0.36) and those receiving cemented implants (p < 0.57). There was no difference in the thirty-day mortality rate for revision as compared with primary hip arthroplasty (p < 0.92).

Conclusions: Factors that are associated with an increased risk of mortality within thirty days after elective hip arthroplasty include an older age, male gender, and a history of cardiorespiratory disease. There has been a significant decline in the thirty-day mortality rate after elective hip arthroplasty in the last decade (p < 0.0002); during the 1990s, the overall rate at our institution was 0.15% (twenty-three of 14,989).

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Total hip arthroplasty is one of the most reliable and cost-effective surgical procedures currently being performed^1-4. Although hip surgery is generally regarded as safe, it is associated with a risk of mortality^5-9. Considerable variation exists in the reported data on perioperative mortality^3,10-12. Even though the overall mortality rate is low, a substantial number of deaths occur in association with more than 300,000 hip replacements that are performed in the United States annually. Identification of risk factors that are associated with perioperative mortality after elective hip surgery may allow clinicians to adjust the perioperative management of patients who are at greatest risk.

We recently reported on the risk factors associated with intraoperative mortality during hip surgery⁷. An underlying diagnosis of fracture (particularly pathological fracture and intertrochanteric fracture), use of cemented implants, older age, and preexisting cardiovascular disease were identified as risk factors for intraoperative death. A major risk factor for intraoperative death in that study was a preoperative diagnosis of fracture. These data suggest that patients undergoing hip arthroplasty after fracture comprise a different population than those undergoing elective total hip surgery, and hence a separate review of the factors influencing perioperative mortality in the latter group seemed warranted. The purpose of the present study was to evaluate the thirty-day mortality after elective total hip surgery with regard to age, gender, implant type, and fixation mode in an effort to identify the significant risk factors for perioperative death.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1:: Graph depicting the thirty-day mortality after elective hip arthroplasty, expressed as a percentage of the total number of such procedures performed during each year of the study.

View Larger

TABLE I: Perioperative Death Rates for Each Suspected Risk Factor

Risk Factor	Total No. of Patients	No. of Patients Who Died (%)	P Value
Age (yr)
70	11,542	51 (0.44)	<0.0002
<70	19,172	39 (0.20)
Gender
Male	14,381	64 (0.45)	<0.0001
Female	16,333	26 (0.16)
Revision
No	23,343	68 (0.29)	?0.92
Yes	7371	22 (0.30)
Cement
No	3669	?9 (0.25)	<0.57
Yes	27,045	81 (0.30)
Diagnosis
Osteoarthritis	15,413	82 (0.53)	<0.36
Rheumatoid arthritis	1072	?8 (0.75)

Add to My JBJS

TABLE I: Perioperative Death Rates for Each Suspected Risk Factor

Risk Factor	Total No. of Patients	No. of Patients Who Died (%)	P Value
Age (yr)
70	11,542	51 (0.44)	<0.0002
<70	19,172	39 (0.20)
Gender
Male	14,381	64 (0.45)	<0.0001
Female	16,333	26 (0.16)
Revision
No	23,343	68 (0.29)	?0.92
Yes	7371	22 (0.30)
Cement
No	3669	?9 (0.25)	<0.57
Yes	27,045	81 (0.30)
Diagnosis
Osteoarthritis	15,413	82 (0.53)	<0.36
Rheumatoid arthritis	1072	?8 (0.75)

View Larger

TABLE II: Cause of Death

*The number of patients who had an autopsy is given in parentheses. †This group included a patient who died intraoperatively and was found to have multiple fat, marrow, and polymethylmethacrylate fragments in the lungs, kidneys, and brain as well as evidence of cardiac ischemia. ‡Other causes of death included hepatic failure, pancreatitis, renal failure, pseudomembranous colitis, and chemical peritonitis from a ruptured gallbladder in one patient each.

Cause of Death	Presumed Cause of Death* (no. of patients)	Cause of Death as Identified at Autopsy (no. of patients)
Myocardial infarction	21 (10)	10
Cardiorespiratory arrest	21 (5)	?3
Pulmonary embolism	14 (12)	14
Multiple system failure	?6 (3)	?3
Sepsis	?5 (5)	?5
Pneumonia	?2 (2)	?2
Gastrointestinal bleeding	?2	—
Aneurysm rupture	?2 (1)	?1
Arrhythmia	?3	—
Congestive cardiac failure	?2	—
Cerebrovascular accident	?1 (1)	?1
Multiple causes†	?6 (1)	?1
Other‡	?5 (1)	?1
Total	90 (41)	41

Add to My JBJS

TABLE II: Cause of Death

Cause of Death	Presumed Cause of Death* (no. of patients)	Cause of Death as Identified at Autopsy (no. of patients)
Myocardial infarction	21 (10)	10
Cardiorespiratory arrest	21 (5)	?3
Pulmonary embolism	14 (12)	14
Multiple system failure	?6 (3)	?3
Sepsis	?5 (5)	?5
Pneumonia	?2 (2)	?2
Gastrointestinal bleeding	?2	—
Aneurysm rupture	?2 (1)	?1
Arrhythmia	?3	—
Congestive cardiac failure	?2	—
Cerebrovascular accident	?1 (1)	?1
Multiple causes†	?6 (1)	?1
Other‡	?5 (1)	?1
Total	90 (41)	41

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

A computer-assisted review of the records of all 38,488 patients who had undergone any type of hip arthroplasty at our institution from 1969 to 1997 was conducted to identify patients who had died within thirty days after the operation. It was determined that 7774 of the procedures had been performed because of a fracture-related diagnosis, leaving 30,714 elective arthroplasties for review.

Patients who did not return for scheduled follow-up appointments at six to twelve weeks or at one or two years were contacted at each time-interval by means of a mailed questionnaire and then with a telephone call, if necessary. These repeated efforts to obtain follow-up information on all patients at a number of time-points gives us great confidence that any patients who may have died during the first thirty days after hip arthroplasty were identified within two years after the operation.

Ninety patients died after elective hip arthroplasty during the study period. The records of these patients were reviewed with regard to age, gender, preoperative diagnosis, medical and surgical history, the use of cement, the type of anesthetic agents used, the course of anesthesia, intraoperative and postoperative events, and signs and symptoms associated with death. The records were specifically examined with regard to cardiac rhythm abnormalities, arterial blood-gas data, the chemical analysis of blood samples, the findings on chest radiographs, and the results of any other diagnostic studies performed during or after hip arthroplasty that could shed light on the cause of death. Pathology records were reviewed to identify the findings at autopsy, if performed, and the autopsy findings were compared with the presumed cause of death as recorded prior to autopsy. The data on patients who died were compared with the data collected from the entire total-joint registry in order to calculate the thirty-day mortality for specific patient subgroups as well as for the entire series of patients who had undergone elective hip arthroplasty.

Statistical Analysis

Individual risk factors were analyzed with the Fisher exact test. Continuous risk factors were analyzed with a two-sample t test. A 95% confidence level was used for all tests.

Results

Introduction | Materials and Methods | Results | Discussion | References

Demographic Data

The ninety patients who died included sixty-four men and twenty-six women. The indication for hip arthroplasty was osteoarthritis in eighty-two patients and rheumatoid arthritis in eight. Sixty-eight patients died after primary hip arthroplasty, and twenty-two patients died after revision arthroplasty. Patients who underwent revision hip arthroplasty were slightly younger than those who underwent primary hip arthroplasty (mean age, sixty-two years compared with sixty-four years; p < 0.08). Patients who received cemented components were older than those who received uncemented components (mean age, sixty-four years compared with fifty-eight years; p < 0.005). Thirteen of the patients who died had had a previous total hip arthroplasty on the contralateral side.

Medical History

The American Society of Anesthesiologists score was introduced at our institution in 1983, and hence the score was not available for 6202 patients; however, it was calculated retrospectively for those who died before 1983. The mean score for the ninety patients who died (2.8; range, 2 to 4) was significantly higher than that for the 24,512 patients who did not die and for whom the score was available (2.1; range, 1 to 4) (p < 0.0001). The mean score for men was significantly higher than that for women, both among patients who died (3.3 compared with 2.6; p < 0.05) and among patients who did not die (2.5 compared with 1.9; p < 0.05). Of the ninety patients who died, sixty-seven had a history of cardiovascular disease and twenty-one had a history of pulmonary disease. Fifteen patients had both cardiovascular and pulmonary disease, and thirty-nine patients had two or more coexistent cardiovascular conditions. The most common conditions were hypertension (thirty-six patients), coronary artery disease (twenty-nine patients), chronic obstructive airway disease (twenty-six patients), a history of myocardial infarction (twenty-three patients), arrhythmia (twenty patients), and congestive cardiac failure (sixteen patients).

Intraoperative Data

Four patients died intraoperatively, and eighty-six died during the thirty-day postoperative period. Examination of the anesthesia records revealed that a sudden hypotensive event, defined as a drop in systolic pressure from a baseline of >20 mm Hg, was observed in sixty-two of the ninety patients who died. Hypotension occurred during cementing of the femoral component in twenty-three patients, during cementing of the acetabular component in thirteen patients, during relocation of the hip in one patient, and during femoral or acetabular preparation in twenty-five patients. Hypotension was accompanied by cardiac arrhythmia in nineteen patients; four of these patients died intraoperatively despite maximal resuscitative efforts. Vasopressors or inotropic agents were used to treat hypotension during the operation on forty-one of the ninety patients who died.

Mortality Rates

The overall thirty-day mortality rate was 0.29% (ninety of 30,714). However, there was a dramatic and substantial decline in the perioperative mortality rate over the nearly three decades encompassed by this study; specifically, the rate fell from 0.94% (thirty-four of 3614) in the 1970s to 0.27% (thirty-three of 12,111) in the 1980s (p < 0.001) and ultimately to 0.15% (twenty-three of 14,989) in the 1990s (p < 0.0002) (Fig. 1).

The thirty-day mortality rate was evaluated with regard to age, gender, type of surgery (primary or revision), type of fixation (with or without cement), and diagnosis (Table I). The mortality rate among patients who were seventy years old or more (0.44%; fifty-one of 11,542) was significantly higher than that among patients who were less than seventy years old (0.20%; thirty-nine of 19,172) (p < 0.0002). Also, the average age of the patients who died (seventy-two years; range, fifty-six to eighty-nine years) was significantly higher than that of the patients who did not die (sixty-four years; range, eleven to 103 years) (p < 0.007). The rate for male patients (0.45%; sixty-four of 14,381) was significantly higher than that for female patients (0.16%; twenty-six of 16,333) (p < 0.0001). The rate for patients managed with revision arthroplasty (0.30%; twenty-two of 7371) was not significantly different from that for patients managed with primary arthroplasty (0.29%; sixty-eight of 23,343) (p = 0.92). The rate for patients who received a cemented implant (0.30%; eighty-one of 27,045) was slightly, but not significantly, higher than that for patients who received an uncemented implant (0.25; nine of 3669) (p < 0.57). The rate for patients with rheumatoid arthritis (0.75; eight of 1072) was slightly higher than that for patients with osteoarthritis (0.53%; eighty-two of 15,413) (p < 0.36).

Analysis of Risk Factors

Univariate analysis revealed that several of the potential risk factors were associated with a significant increase in the thirty-day mortality rate. These included increased age (more than seventy years) (p < 0.0002), male gender (p < 0.0001), a history of cardiorespiratory disease (p < 0.0001), and the decade in which the operation was performed. Factors that were not significant included the use of cement (p < 0.57), the underlying diagnosis (p < 0.36), the side of the operation (p < 0.82), and the type of arthroplasty (primary or revision) (p < 0.92). Analysis of risk factors was also carried out for various clinical subgroups. The relative risk of death was found to be 10.1 times greater (95% confidence interval, 6.6 to 15.6) for patients who were seventy years old or more, 5.2 times greater (95% confidence interval, 2.6 to 10.5) for patients with an underlying diagnosis of cardiorespiratory disease, and 3.7 times greater (95% confidence interval, 2.3 to 6.0) for men.

Cause and Time of Death

An autopsy was performed on forty-one of the ninety patients who died. The major findings were limited to the lungs and heart. There was a close agreement between the presumed cause of death and the autopsy findings for the majority of the patients (Table II). Twenty-one patients were presumed to have died of sudden cardiorespiratory arrest; almost all of these patients died within forty-eight hours after the index operation. The death of many of these patients could not be attributed solely to a cardiac or pulmonary cause but often was thought to be due to a combination of factors. Thirty-nine deaths occurred during or within the first twenty-four hours after the operation, and seventy-nine (88%) of the ninety deaths occurred in the hospital.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

An important finding of our study is the increased perioperative mortality seen with increasing age^4,13. Elderly patients are more likely to have preexisting comorbidities and may be less able to respond physiologically to the hemodynamic insults of surgery.

The other striking finding of this study was the presence of severe underlying cardiovascular and pulmonary disease in the majority of patients who died. The exact prevalence of cardiorespiratory conditions for patients in our entire total joint-registry population is unknown. However, the mean American Society of Anesthesiologists score was higher for those who died compared with those who did not die. Furthermore, the mean score was higher for male patients, perhaps partially explaining the higher mortality rate among men⁴.

This study also identified a dramatic decline in the perioperative mortality after elective hip arthroplasty over the years (Fig. 1). The reasons for this decline are likely multifactorial and may relate to the combined effects of changes in surgical techniques, anesthesia management, and perioperative medical management that have been made during this time^12,14,15.

The majority of our patients who died exhibited cardiorespiratory disturbances during the index operation. Often, these disturbances took place during cementing of a component or during other steps in the arthroplasty procedure that are known to be associated with embolic showers¹⁶. All of the intraoperative deaths that have occurred during hip arthroplasty at our institution have been associated with events that are known to produce embolization of marrow contents. It is possible that some of the postoperative deaths were also due, in part, to the late sequelae of intraoperative embolization of marrow and fat elements from the bone, perhaps explaining the relatively high number of deaths that occurred early after the operation.

In conclusion, factors that are associated with increased perioperative mortality after elective hip arthroplasty include an older age, male gender, and a history of cardiorespiratory disease. Awareness of these factors can alert the surgeon to patients who are potentially at risk and thereby facilitate management, particularly during the preoperative and perioperative periods. The identification of patients who are at risk also allows for alterations in anesthetic and operative techniques as well as in implant selection in order to minimize the possibilty of adverse events. These findings also should help to determine the focus of future research efforts aimed at further reducing this uncommon, yet devastating, complication of elective hip arthroplasty.

References

Introduction | Materials and Methods | Results | Discussion | References

Melton LJ 3rd. Epidemiology of fractures. In: Riggs BL, Melton LJ 3rd, editors. Osteoporosis: etiology, diagnosis, and management. New York: Raven Press; 1988. p 133-54.

Surin VV,Sundholm K. Survival of patients and prostheses after total hip arthroplasty. Clin Orthop,1983;177: 148-53. 177148 1983 [PubMed]

Taylor HD, Dennis DA,Crane HS. Relationship between mortality rates and hospital patient volume for Medicare patients undergoing major orthopaedic surgery of the hip, knee, spine, and femur. J Arthroplasty,1997;12: 235-42. 12235 1997 [PubMed]

Whittle J, Steinberg EP, Anderson GF, Herbert R,Hochberg MC. Mortality after elective total hip arthroplasty in elderly Americans. Age, gender, and indication for surgery predict survival. Clin Orthop,1993;295: 119-26. 295119 1993 [PubMed]

Coventry MB, Beckenbaugh RD, Nolan DR,Ilstrup DM. 2,012 total hip arthroplasties. A study of postoperative course and early complications. J Bone Joint Surg Am,1974;56: 273-84. 56273 1974 [PubMed]

Mullen JO,Mullen NL. Hip fracture mortality. A prospective, multifactorial study to predict and minimize death risk. Clin Orthop,1992;280: 214-22. 280214 1992 [PubMed]

Parvizi J, Holiday AD, Ereth MH,Lewallen DG. Sudden death during primary hip arthroplasty. Clin Orthop,1999;369: 39-48. 36939 1999 [PubMed]

Patterson BM, Healey JH, Cornell CN,Sharrock NE. Cardiac arrest during hip arthroplasty with a cemented long-stem component. A report of seven cases. J Bone Joint Surg Am,1991;73: 271-7. 73271 1991 [PubMed]

Schoning B, Schulitz KP,Pfluger T. Statistical analysis of perioperative and postoperative mortality of patients with prosthetic replacement of the hip joint. Arch Orthop Trauma Surg,1980;97: 21-6. 9721 1980 [PubMed]

Colwell CW Jr, Hinson J, McCutchen J, Paulson R,Collis D. Enoxaparin vs. warfarin: hospital management for DVT prevention [abstract]. J Arthroplasty,1997;12: 229. 12229 1997

Dearborn JT,Harris WH. Postoperative mortality after total hip arthroplasty. An analysis of deaths after two thousand seven hundred and thirty-six procedures. J Bone Joint Surg Am,1998;80: 1291-4. 801291 1998 [PubMed]

Sharrock NE, Cazan MG, Hargett MJ, Williams-Russo P,Wilson PD. Changes in mortality after total hip and knee arthroplasty over a ten-year period. Anesth Analg,1995;80: 242-8. 80242 1995 [PubMed]

White RH, McCurdy SA,Marder RA. Early morbidity after total hip replacement: rheumatoid arthritis versus osteoarthritis. J Gen Intern Med,1990;5: 304-9. 5304 1990 [PubMed]

Babu SC, Sharma PV, Raciti A, Mayr CH Jr, Elrabie NA, Clauss RH, Stahl WM,Del Guercio LR. Monitor-guided responses. Operability with safety is increased in patients with peripheral vascular diseases. Arch Surg,1980;115: 1384-6. 1151384 1980 [PubMed]

Berlauk JF, Abrams JH, Gilmour IJ, O’Connor SR, Knighton DR,Cerra FB. Preoperative optimization of cardiovascular hemodynamics improves outcome in peripheral vascular surgery. A prospective, randomized clinical trial. Ann Surg,1991;214: 289-99. 214289 1991 [PubMed]

Ereth MH, Weber JG, Abel MD, Lennon RL, Lewallen DG, Ilstrup DM,Rehder K. Cemented versus noncemented total hip arthroplasty—embolism, hemodynamics, and intrapulmonary shunting. Mayo Clin Proc,1992;67: 1066-74. 671066 1992 [PubMed]

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1:: Graph depicting the thirty-day mortality after elective hip arthroplasty, expressed as a percentage of the total number of such procedures performed during each year of the study.

View Larger

TABLE I: Perioperative Death Rates for Each Suspected Risk Factor

Risk Factor	Total No. of Patients	No. of Patients Who Died (%)	P Value
Age (yr)
70	11,542	51 (0.44)	<0.0002
<70	19,172	39 (0.20)
Gender
Male	14,381	64 (0.45)	<0.0001
Female	16,333	26 (0.16)
Revision
No	23,343	68 (0.29)	?0.92
Yes	7371	22 (0.30)
Cement
No	3669	?9 (0.25)	<0.57
Yes	27,045	81 (0.30)
Diagnosis
Osteoarthritis	15,413	82 (0.53)	<0.36
Rheumatoid arthritis	1072	?8 (0.75)

Add to My JBJS

TABLE I: Perioperative Death Rates for Each Suspected Risk Factor

Risk Factor	Total No. of Patients	No. of Patients Who Died (%)	P Value
Age (yr)
70	11,542	51 (0.44)	<0.0002
<70	19,172	39 (0.20)
Gender
Male	14,381	64 (0.45)	<0.0001
Female	16,333	26 (0.16)
Revision
No	23,343	68 (0.29)	?0.92
Yes	7371	22 (0.30)
Cement
No	3669	?9 (0.25)	<0.57
Yes	27,045	81 (0.30)
Diagnosis
Osteoarthritis	15,413	82 (0.53)	<0.36
Rheumatoid arthritis	1072	?8 (0.75)

View Larger

TABLE II: Cause of Death

Cause of Death	Presumed Cause of Death* (no. of patients)	Cause of Death as Identified at Autopsy (no. of patients)
Myocardial infarction	21 (10)	10
Cardiorespiratory arrest	21 (5)	?3
Pulmonary embolism	14 (12)	14
Multiple system failure	?6 (3)	?3
Sepsis	?5 (5)	?5
Pneumonia	?2 (2)	?2
Gastrointestinal bleeding	?2	—
Aneurysm rupture	?2 (1)	?1
Arrhythmia	?3	—
Congestive cardiac failure	?2	—
Cerebrovascular accident	?1 (1)	?1
Multiple causes†	?6 (1)	?1
Other‡	?5 (1)	?1
Total	90 (41)	41

Add to My JBJS

TABLE II: Cause of Death

Cause of Death	Presumed Cause of Death* (no. of patients)	Cause of Death as Identified at Autopsy (no. of patients)
Myocardial infarction	21 (10)	10
Cardiorespiratory arrest	21 (5)	?3
Pulmonary embolism	14 (12)	14
Multiple system failure	?6 (3)	?3
Sepsis	?5 (5)	?5
Pneumonia	?2 (2)	?2
Gastrointestinal bleeding	?2	—
Aneurysm rupture	?2 (1)	?1
Arrhythmia	?3	—
Congestive cardiac failure	?2	—
Cerebrovascular accident	?1 (1)	?1
Multiple causes†	?6 (1)	?1
Other‡	?5 (1)	?1
Total	90 (41)	41