Total joint arthroplasty represents one of the greatest advances in modern
orthopaedic surgery and continues to be one of the safest and the most
effective surgical procedures. Nevertheless, hip and knee arthroplasty have
been associated with several serious potential complications and occasionally
even with
death1-4.
Recent improvements in surgical and anesthesia techniques have contributed
to a marked reduction in mortality after elective joint
arthroplasty3,4.
However, because modern medical care results in longer life expectancy in
general and in better survival of patients with cardiac and other systemic
illnesses in particular, joint arthroplasty is being performed in sicker
patients. Hence, it is plausible that orthopaedic surgeons may observe a
higher prevalence of medical complications in their patients who have elective
joint arthroplasty.
The latter, if true, would imply that better preoperative medical
optimization and stringent, perhaps longer, postoperative surveillance of
these patients are required. In contrast, with the surge in minimally invasive
joint
arthroplasty5-13,
the decline in reimbursement, and patients' expectation for faster recovery, a
worrisome trend for shorter hospital stays has been emerging.
This prospective study was designed to elucidate the incidence, timing, and
the severity of complications that may occur following elective primary
unilateral total joint arthroplasty in the lower extremity. Furthermore, we
sought to specifically explore the nature of these complications with the
intention of identifying patients who may be at risk.
Demographic Data
All patients undergoing elective primary or revision total hip or knee
arthroplasty at our institution from July 2004 to July 2005 were recruited
into the study. Institutional review board approval was obtained for the
study. During this period, a total of 2391 hip and knee reconstructive
procedures had been performed in 2183 patients, and 2048 of the procedures
were total joint arthroplasties done in 1842 patients. This included
1056 primary total hip arthroplasties in 1011 patients, 992 primary total knee
arthroplasties in 831 patients, 198 revision total hip arthroplasties, and 120
revision total knee arthroplasties. The primary joint arthroplasties had been
performed in 732 male patients with a mean age of sixty-two years (range,
fifteen to ninety-one years) and 1110 female patients with a mean age of
sixty-six years (range, seventeen to ninety-seven years). This study includes
only patients undergoing primary unilateral total joint arthroplasty.
Study Plan
Preoperative Evaluation
All patients who underwent elective joint arthroplasty at our institution
completed preoperative evaluation forms that captured detailed data including
demographic information, past medical and surgical history, medication
history, and the functional status.
In-Hospital Course
A standardized and detailed form intended to capture all medical and
orthopaedic complications was devised. This form was completed on a daily
basis for every patient who underwent elective joint arthroplasty. The
hospital course of the patients was followed very closely, and any
complications identified by the treating internist or evident on a laboratory
test or other investigations were recorded. The details of any medical or
surgical intervention were also noted.
Course After Discharge
During the first postoperative visit, occurring usually at six weeks, the
postoperative evaluation forms completed by the patient were collected. A form
for capturing any complications that may have happened following discharge was
also completed by the examining physician. The circumstances leading to the
complications, if any, and the details of the therapeutic intervention for
each complication were recorded. As part of routine surveillance, patients who
did not return for scheduled follow-up appointments were contacted by means of
a detailed questionnaire sent by mail and then by a telephone call, if
necessary. The mean duration of follow-up for the patients in this cohort was
ninety-two days (range, sixty to 420 days).
Surgical Data
Regional anesthesia, unless it was contraindicated, was used for all
patients who had elective arthroplasty. All hip arthroplasties were performed
through an anterolateral approach with the patient in the supine position. All
total knee arthroplasties were performed through a medial parapatellar
arthrotomy. All operations were performed by or under the close supervision of
one of five arthroplasty surgeons. Joint replacements were carried out in a
laminar flow room with all members of the surgical team wearing body-exhaust
suits during surgery. All patients in this cohort received uncemented total
hip replacements. All patients who had a primary hip arthroplasty were allowed
to walk postoperatively using a walking aid with weight-bearing as
tolerated.
Postoperative Care
All patients were followed closely by an internist. Patients deemed to be
at high risk of cardiovascular complications postoperatively were routinely
admitted to the intensive care unit or a step-down unit with continuous
electrocardiogram monitoring and dedicated nursing care. For other patients,
monitoring of the respiratory rate, heart rate, blood pressure, and pulse
oximetry was carried out at frequent intervals by the nursing staff. Any
changes in these parameters that were deemed to be clinically important were
investigated. Daily laboratory tests, including complete blood-cell count,
renal function tests, blood chemistry, prothrombin time, and international
normalized ratio, were performed. Additional tests, such as liver function
tests, troponin levels, chest radiographs, electrocardiograms, and
computerized tomography, were also ordered as deemed necessary for some
patients.
Thromboembolic prophylaxis, namely low-dose Coumadin (warfarin) (with a
goal of an international normalized ratio in the range of 1.5 to 2.0) for six
weeks, was administered to all patients unless contraindicated. Intravenous
antibiotics, usually a first-generation cephalosporin, were administered
preoperatively and for twenty-four hours following the surgery.
Definition of Complications
The complications were categorized into systemic and local. Furthermore,
the complications were subcategorized as major systemic if they were
deemed to be life-threatening and required complex medical intervention, such
as transfer to the intensive care unit, or as major local if the
complication necessitated additional surgical intervention or was deemed to
result in temporary or permanent functional impairment. Minor complications
were those that resulted in a prolonged hospital stay, necessitated additional
observation such as monitoring of the wound, or required additional medical
treatment such as blood transfusion or antibiotics.
Myocardial infarction was determined on the basis of an increase in the
troponin level and/or electrocardiographic changes. Acute renal failure was
defined as an abrupt decline in renal function parameters (increasing
creatinine) with reduced urine output. Pulmonary embolus was diagnosed on the
basis of the presence of emboli in the pulmonary vasculature as determined by
multidetector computerized tomography. Anemia was thought to exist if the
postoperative hemoglobin level of the patient, regardless of the preoperative
hemoglobin level, had declined to <10 mg/dL. Persistent wound drainage was
defined as any degree of drainage from the wound more than forty-eight hours
after the index surgery. Deep venous thrombosis was diagnosed with use of
Doppler ultrasound and was categorized into distal and proximal clots.
Statistical Analysis
Descriptive statistics and the Fisher exact test were used for analysis of
categorical data. Continuous data were analyzed with use of descriptive
statistics and the non-paired t test. Univariate regression analysis was
performed on all risk factors for complications. Multiple logistic regression
analysis was performed on all variables that were significant in univariate
analyses. For the univariate analysis, a p value of < 0.1 was considered
significant. For all other tests, a p value of < 0.05 was considered
significant. All analyses were performed with use of SPSS software (version
13; SPSS, Chicago, Illinois).
Cause and Time of Death
One patient (0.06%) in the cohort died. The patient had had aspiration
pneumonia develop three days after a primary knee arthroplasty and died on the
twenty-third postoperative day while in the intensive care unit. Postmortem
examination revealed the final cause of death to be cerebral edema secondary
to aspiration pneumonia.
Incidence of Medical Complications
There were a total of 104 life-threatening systemic complications
(Table I). These included a
non-fatal cardiac arrest in one patient, tachyarrhythmia in thirty-three
patients, congestive heart failure or pulmonary edema in ten patients, and
myocardial infarction in six patients. Four patients had development of severe
hypotension that necessitated vasopressor administration and transfer to a
monitored bed. The cause of the hypotension in all four patients could not be
elicited with certainty, but it was deemed to be cardiac in origin. There were
twenty-five confirmed pulmonary emboli. Other complications included acute
renal failure in fourteen patients, stroke or a transient ischemic attack in
six patients, bowel obstruction in three patients, and pneumothorax in one
patient. A total of 353 minor systemic complications, mostly related to
anemia, also occurred in this cohort.
Incidence of Orthopaedic Complications
There were seventeen major and ninety-nine minor orthopaedic complications.
Surgical intervention to address the complications was required in eleven
patients and included leg fasciotomy following vascular injury (two patients),
embolectomy (two patients), open reduction and internal fixation of
periprosthetic fractures (two patients), and operative incision and drainage
(five patients).
Timing of Complications
Ninety-four of the life-threatening medical complications occurred within
four days of the index surgery (Fig.
1). The mean length of the hospital stay for the entire cohort was
3.9 days (range, one to thirty-six days) and 4.7 days (range, two to
thirty-six days) for the patients who had complications (p < 0.001). There
was one major complication (stroke) that occurred following discharge of the
patient from the hospital.
Risk Factors for Major Complications
A number of factors were analyzed with respect to their potential influence
on the development of major medical complications
(Table II). Univariate analysis
identified older age; a higher body-mass index; general anesthesia; and
medical comorbidi-ties, as measured by the American Society of Anesthesiology
(ASA) score, to be important predictors of postoperative major complications.
When these factors were subjected to multiple logistic regression analysis,
age (odds ratio, 1.21; 95% confidence interval, 1.02 to 1.19), body mass index
(odds ratio, 1.01; 95% confidence interval, 0.99 to 1.09), and ASA score (odds
ratio, 0.76; 95% confidence interval, 0.55 to 1.2) were found to be important
predictors of major complications.
Readmissions
There were a total of ten readmissions in the series. The reasons for
readmission included wound-related problems in six patients. The reasons for
readmission for medical complications included deep-vein thrombosis in four
patients and stroke in one patient. One of the readmitted patients experienced
two complications, resulting in a total of eleven complications in these
patients.
Substantial and parallel advances in the medical fields over the recent
years have enabled a large number of patients with severe illnesses to enjoy
longer life expectancy and to seek orthopaedic care for their degenerative
joints. Hence, total joint arthroplasty is currently being offered to some
patients with serious
comorbidities14-16
who may have been deemed inappropriate candidates during the early years of
joint replacement. Despite the availability of this surgical procedure to sick
and frail patients, the mortality and morbidity following total joint
arthroplasty remains very
low4. Deaths and
complications following joint arthroplasty, however, can and do
occur4.
In recent years, minimally invasive surgery has become more
popular6,8,13,17,18.
The advocates of minimally invasive surgery have cited reduced intraoperative
blood loss; less peri-operative pain; faster recovery; better incision
cosmesis; and, more importantly, a shorter hospital stay as potential
benefits6-8,13,19.
The mean length of stay following minimally invasive total hip arthroplasty
was reported to be 1.9 days compared with 3.5 days for a standard hip
arthroplasty6.
Berger and Duwelius proposed a "same day" discharge pathway for
patients undergoing two-incision total hip
arthroplasty7.
The minimally invasive arthroplasty, with the exception of some inherent
differences in soft-tissue dissection, still involves surgical violation of
the medullary canal with the potential for fat and marrow
embolization14,20,21.
Furthermore, patients require the same degree of anesthesia as that needed for
conventional joint replacement to allow the surgeon to perform the same bone
cuts and insert the same implants as in conventional joint arthroplasty.
Hence, any complications associated with a standard joint arthroplasty can
conceivably occur after minimally invasive joint replacement. This raises a
critical and worrisome concern: what influence does a shorter hospital stay
have on the management of life-threatening complications if they were to occur
outside the hospital?
This study, which was designed with the intention of identifying
in-hospital complications that may occur after elective joint arthroplasty,
highlights some important findings. First, it demonstrated that a considerable
number of life-threatening medical complications occur in 6% of the patients
who have an elective joint arthroplasty. All of these complications were
deemed to be life-threatening, as the lack of emergent medical intervention in
the form of the administration of drugs and close monitoring could have
resulted in catastrophic consequences. Even if one assumes that emergent
medical care could have been provided for most of these patients if they were
outside the hospital, one cannot overlook the relatively high incidence of
cardiac complications in general and near-fatal arrhythmias (ventricular
tachycardia) in particular that occurred in these patients. On the basis of
the analysis of detailed data, we can confidently say that major
complications, if they occurred outside the hospital, would have resulted in
the death of twenty to twenty-five patients in this cohort. Another important
finding of this study was that for more than half (58%) of the patients who
had a major complication, the complication could not have been predicted on
the basis of the medical history. Hence, strategies such as close surveillance
of so-called at risk patients may not identify other patients who may have
life-threatening complications develop for the first time. Furthermore, all of
these life-threatening complications occurred despite intensive and rigorous
medical evaluation of all patients in this cohort. Finally, the present study
demonstrated that the majority (>90%) of the fatal and near-fatal
complications occurred during the early days and within the expected typical
four-day hospital stay of a standard elective joint replacement
(Fig. 1).
The incidence of in-hospital mortality and major complications such as
myocardial infarction in our cohort is similar to that in previous
reports10,11.
Furthermore, the risk factors identified in our study, namely old age and the
history of coexistent diseases, also agree with the findings of previous
studies10. An
additional factor, increased body mass index, was identified as a possible
etiological factor for an increased rate of complications. Although one may
have intuitively selected any of the aforementioned as possible predisposing
factors to systemic complications after total joint arthroplasty, it is
important to recall that not all patients with near-fatal complications
possessed these risk factors. As far as we know, none of these risk factors
have been cited as contraindications for minimally invasive total hip
arthroplasty. We deliberately excluded patients who had a revision or
bilateral joint arthroplasty since, by strict criteria, they may not be
considered as candidates for minimally invasive joint arthroplasty. Hence, any
of the complications that were encountered in this cohort could occur in
patients undergoing minimally invasive joint replacement.
We invested extensive efforts to ensure that every complication occurring
in the hospital and following discharge was captured. Despite the prospective
nature of this study, it is possible, although unlikely, that some
complications may have escaped detection particularly if they occurred
following discharge. The latter, if true, would imply that the actual
incidence of complications is higher than what is reported. We believe that
the incidence of complications reported in the present study is likely to be
affected by other factors. One important fact to consider is that this study
was performed in a single, high-volume institution where standardized
protocols are in place and patients receive a thorough preoperative medical
evaluation and are followed diligently by internists postoperatively. Hence,
it is plausible that the incidence of complications would be higher in centers
if the aforementioned protocols were not in place. Conversely, the
implementation of these protocols, in particular the routine measurement of
pulse oximetry and the availability of modern imaging modalities such as
multidetector computed tomography scanning, may have resulted in the detection
of complications that might have otherwise gone unnoticed. We were also
stringent with the definition of some complications such as anemia and wound
drainage. The majority of the patients diagnosed as having anemia in this
study did not receive a blood transfusion and were asymptomatic. A large
number of patients with chronic disease who had a low preoperative hemoglobin
level, and despite experiencing a small decrease in the hemoglobin level, were
categorized as anemic. Similarly, a large number of patients in this cohort
who were classified as having a draining wound would not have been classified
as such in other centers. We had selected any drainage from the wound beyond
forty-eight hours as the cut-off beyond which drainage was considered
abnormal.
Our findings raise concern and caution against the current strategies to
reduce hospital length of stay after total reconstruction of a major joint in
the lower extremity. ?