We conducted a retrospective review of the cases of patients who underwent
either a primary or revision total hip or knee arthroplasty performed by the
senior author (K.L.G.) from October 1995 to January 2002. Of the 1170 patients
who had total hip or knee arthroplasty performed during the study period,
forty patients who had abdominal symptoms in the acute postoperative period
(within ten days) that were severe enough to warrant abdominal radiographs
were identified. The radiographs from these forty patients were analyzed for
the presence of a normal gas pattern, postoperative ileus, or Ogilvie
syndrome. Postoperative ileus was diagnosed radiographically by the presence
of a gas pattern involving the entire bowel with no preferential large bowel
dilatation (Fig. 1). The
diagnosis of Ogilvie syndrome was based on the radiographic presence of
preferential dilatation of the large bowel (most commonly the cecum and
transverse colon) in the absence of mechanical obstruction with minimal to no
gas in the small bowel (Fig.
2). None of the forty patients had a normal gas pattern;
twenty-two of these patients were diagnosed as having postoperative ileus, and
the remaining eighteen patients were diagnosed as having Ogilvie syndrome.
These eighteen patients are the focus of the present study. Institutional
review board approval was obtained for this series, and all of the patients
provided informed consent.
The medical records of the study group were reviewed for patient age,
gender, type of arthroplasty performed, preoperative medical conditions,
preoperative and postoperative medications, route of anesthesia, total
anesthesia time, grade according to the system of the American Society of
Anesthesiologists17,
position of the patient during the procedure, type of narcotic used in the
acute postoperative period and the method of administration
(patient-controlled analgesia or administration by a nurse as needed),
promotility agent(s) given in the postoperative period, postoperative
electrolyte abnormalities, time of onset of symptoms, type of symptoms, type
of treatment, duration of symptoms, and length of hospitalization.
The data on the study group were compared with those for a control group,
which consisted of 160 randomly selected patients who underwent total hip or
knee arthroplasty during the same study period. This was a procedure-matched
control group consisting of forty-five patients who had had a primary total
hip arthroplasty, thirty-seven who had had a primary total knee arthroplasty,
forty-four who had had a revision total hip arthroplasty, and thirty-four who
had had a revision total knee arthroplasty. The study and control groups were
compared with respect to the differences in the length of postoperative
hospitalization, usage of patient-controlled analgesia in the postoperative
period, patient age, grade according to the system of the American Society of
Anesthesiologists, and total anesthesia time. Also, both the study and control
groups were assessed for the presence of previously identified risk factors
for Ogilvie syndrome associated with medical illnesses and non-arthroplasty
surgical
procedures8,9,12,13,18-20.
In order for a patient to be considered at risk, he or she had to have at
least one of the following: a preoperative diagnosis of diabetes mellitus,
hypothyroidism, electrolyte abnormalities, uremia, bowel disease, previous
abdominal surgery, or preoperative use of medications that are known to
decrease gastrointestinal motility (narcotic pain medications, phenothiazines,
tricyclic antidepressants, calcium channel blockers, H2 receptor
antagonists, and
anticholinergics)9,18,20
(see Appendix).
The treatment algorithm used for patients presenting with signs and
symptoms of abdominal obstruction after total hip or total knee arthroplasty
is shown in Figure 3. After the
patient became symptomatic, oral intake was restricted and abdominal
radiographs were made. Narcotics were discontinued, and walking was increased
as tolerated. A nasogastric tube was placed if the patient complained of
nausea and vomiting in addition to abdominal distension. Serial plain supine
abdominal radiographs were assessed for signs of resolution or worsening. The
patients who did not respond to these preliminary measures received a
gastroenterology or general surgery consultation. The use of a promotility
agent (neostigmine or metoclopramide) and the need for a decompressive
colonoscopy were determined by the consulting gastroenterologist or general
surgeon. The resolution of symptoms was marked by the resumption of diet
(tolerance of a clear to full liquid diet). Statistical analysis was conducted
with use of the Wilcoxon and Fisher exact test. Statistical significance was
considered to be a p value of <0.05.
Incidence
The diagnosis of Ogilvie syndrome was made in eighteen (1.5%) of the 1170
patients who had a total hip or knee arthroplasty performed during the study
period (Table I). The incidence
of Ogilvie syndrome following any type of total hip arthroplasty (primary and
revision) during the study period was 1.6% (eleven of 708 patients). Five
(1.2%) of the 433 primary total hip arthroplasties were associated with the
development of Ogilvie syndrome compared with six (2.2%) of 275 revision total
hip arthroplasties. The difference was not significant (p < 0.20), with the
numbers available.
With respect to the knee arthroplasties (primary and revision), seven
(1.5%) of the 462 patients had Ogilvie syndrome develop. Five (1.3%) of the
380 patients who had a primary total knee arthroplasty performed during the
study period had Ogilvie syndrome develop compared with two (2.4%) of
eighty-two patients who underwent revision total knee arthroplasty. The
difference was not significant (p < 0.20), with the numbers available.
Age and Gender
Of the eighteen patients who had Ogilvie syndrome, twelve were men and six
were women, indicating an increased relative risk for men (odds ratio = 3.75,
p < 0.01). The patients with Ogilvie syndrome were significantly older
(median, eighty years; range, forty-six to eighty-six years) than those in the
control group (median, sixty-seven years; range, twenty to ninety years) (p
< 0.001).
Risk Factors
One or more previously documented risk factors for Ogilvie syndrome,
besides age and gender, were identified in seventeen of the eighteen study
patients. Seven patients had previous abdominal surgery, three had
hypothyroidism, three had Crohn disease, two had diabetes mellitus, two had
diverticulosis, one patient had cholelithiasis, one had a history of a
perforated colon, and one had Parkinson disease with a profound decrease in
gastrointestinal motility for which he required multiple promotility agents.
In addition, thirteen of the eighteen patients were on one or more
preoperative medications considered to be a risk factor for the development of
acute colonic pseudo-obstruction.
In contrast, risk factors for acute colonic pseudo-obstruction were found
in fifty-eight (36%) of the 160 patients in the control group
(Table II). The difference
between the Ogilvie syndrome and control groups, with respect to the
prevalence of risk factors, was significant for all arthroplasties (p <
0.0001), for total hip arthroplasties (p < 0.01), and for total knee
arthroplasties (p = 0.012). Furthermore, the preoperative use of medications
at risk for decreasing gastrointestinal motility was significantly more
frequent in the Ogilvie group (thirteen of eighteen patients) compared with
the control group (thirty-nine [24%] of 160 patients) (p < 0.001).
Symptoms
The most common presenting symptom was abdominal distension, which occurred
in all eighteen patients. Abdominal discomfort was present in sixteen of the
eighteen patients. Fifteen patients had nausea, and thirteen had vomiting.
Flatus or a bowel movement had occurred prior to the onset of symptoms in
sixteen and fourteen patients, respectively.
Onset of Symptoms and Patient-Controlled Analgesia
The onset of symptoms occurred at an average of 3.4 days (range, one to ten
days) after the day of surgery. Symptoms developed at a median of 2.2 days
(range, two to ten days) postoperatively in the ten patients who used
patient-controlled analgesia compared with a median of 4.6 days (range, one to
seven days) in the eight patients who did not use patient-controlled
analgesia. Although a trend was noted (p = 0.067), this difference was not
found to be significant on the basis of the number of study patients.
Patient-controlled analgesia was used in ten of the eighteen patients in the
study group compared with 133 (85%) of 157 subjects in the control group (data
were not available for three subjects); the difference was significant (p <
0.01).
Anesthesia
General anesthesia was used for all of the patients in the study group, and
four of them received an epidural block before the induction of general
anesthesia. The total time under anesthesia (defined as the time from the
first administration of the intravenous agent to the cessation of all
anesthetic medications at the end of the operation) for all procedures was a
median of 210 minutes for the patients who had Ogilvie syndrome develop and a
median of 210 minutes for the control group.
According to the system of the American Society of Anesthesiologists, four
patients were classified as grade 2 and fourteen as grade 3; no patient was
grade 1. Compared with the control group, the study group demonstrated no
significant increase in the incidence of Ogilvie syndrome between the patients
who were grade 3 and those who were grade 2 (p = 0.47), on the basis of the
numbers in this study.
Treatment
Following the treatment protocol as outlined above, fifteen of the eighteen
patients received a gastroenterology or general surgery consultation. The use
of the promotility agent neostigmine was used in only one patient. Seven of
the eighteen patients underwent decompressive colonoscopy at an average of 2.7
days (range, one to four days) after the index arthroplasty. No patient
required a second decompression, and no patient required a laparotomy for
decompression. The resolution of symptoms occurred at an average of 5.4 days
(range, three to eight days) for the patients who underwent decompressive
colonoscopy compared with an average of 8.4 days (range, one to eighteen days)
for those who did not receive this intervention; the difference was not
significant (p = 0.44), with the numbers available. However, a significant
difference was noted between the patients who had a decompressive colonoscopy
and those who did not with respect to the length of hospitalization (p =
0.019). The median length of hospitalization was 12.3 days (range, nine to
twenty-one days) for the patients who received a decompressive colonoscopy
compared with seventeen days (range, eleven to seventy-three days) for those
who did not undergo decompression.
The median length of hospitalization was thirteen days (range, nine to
seventy-three days) for the eighteen study patients and seven days (range, two
to forty-four days) for the control group. The difference was significant (p
< 0.0001).
Complications
Complications in the study group included anemia, which was defined as a
hemoglobin level of <10 (three patients); renal insufficiency (three
patients); urinary tract infection (two patients); and cholecystitis
necessitating cholecystectomy, upper gastrointestinal bleeding, pneumonia with
a pleural effusion, and myocardial infarction (one patient each). There were
no intestinal perforations and no deaths.
Follow-up
Follow-up of the study group was performed in February 2005. Of the
eighteen patients, four had died at eighteen, nineteen, sixty-five, and
seventy-eight months postoperatively. One of the patients who had died had
been diagnosed as having an ileus four years postoperatively. Of the fourteen
surviving patients, one had subsequent development of an intestinal
obstruction (without a preceding surgical intervention) and was treated with a
laparotomy and adhesion release. Two patients had undergone an additional
total joint arthroplasty since the occurrence of Ogilvie syndrome; neither
patient had obstructive symptoms develop in the postoperative period. One
patient reported continued problems with bowel recovery following a subsequent
non-orthopaedic operation. One patient had been diagnosed as having
diverticulitis since the occurrence of Ogilvie syndrome. One patient had
subsequently undergone endoscopy for gastritis. The patient with Parkinson
disease had continued problems with constipation. Five patients had no
additional problems after the Ogilvie syndrome episode, and two patients were
lost to follow-up.
Ogilvie syndrome is a potentially fatal postoperative complication that has
been reported after a variety of surgical procedures, including arthroplasty
in the lower
extremity1-13.
Most of the reports of Ogilvie syndrome that we reviewed consisted of small
heterogeneous groups of patients treated surgically for gastrointestinal or
genitourinary problems, orthopaedic trauma, or arthroplasty in the lower
extremity. Few studies have specifically investigated this complication after
arthroplasty of the hip or knee. Furthermore, to our knowledge, only nineteen
patients with Ogilvie syndrome after total knee replacement have been
reported, with eleven of them described in a single
study7,9,12,19,21.
Despite the relatively infrequent reporting of this complication in the
arthroplasty literature, several risk factors for acute colonic
pseudo-obstruction have been identified; the most prevalent factors include
advanced age, male gender, diabetes mellitus, hypothyroidism, electrolyte
abnormalities, uremia, bowel disease, previous bowel surgery, and use of
maintenance medications known to decrease gastric
motility2,8,9,12,13,21.
The reported incidence of acute colonic pseudo-obstruction after total hip
arthroplasty has ranged from 0.29% to
5.6%8,9,11,12,21.
In the largest study of patients with acute colonic pseudo-obstruction after
total hip arthroplasty that we are aware of, Clarke et
al.8 reported that
the complication developed in thirty of 10,468 patients, for an overall
incidence of 0.29%, although the authors stated that their numbers likely
represent an underestimation of the true rate. The incidence of 1.6% in the
present study is closer to that in the study by Petrisor et
al.12, who reported
that twenty (1.7%) of 1181 patients had this complication develop after total
hip arthroplasty. In contrast to the findings in our series and in the study
by Clarke et al., Petrisor et al. found an increase in the incidence of acute
colonic pseudo-obstruction after revision total hip arthroplasty (5.6%)
compared with that after primary total hip arthroplasty (1.4%); the difference
was significant (p = 0.0043). The rate after total knee arthroplasty in the
current series was seven (1.5%) of 462 patients, which was twice the rate of
0.7% (eleven of 1513 patients) reported by Petrisor et
al.12. In both
reports, no difference was found between the incidence of acute colonic
pseudo-obstruction after primary total knee arthroplasty and that after
revision total knee
arthroplasty12.
One of the goals of this study was to identify risk factors for the
development of Ogilvie syndrome after hip or knee arthroplasty. Clarke et al.
reported an increased risk of acute colonic pseudo-obstruction in patients of
advanced age, male gender, and those who underwent nonelective hip
arthroplasty (that is, following trauma). Petrisor et al., in the only prior
study we are aware of that includes a control group, found bilateral total
knee arthroplasty (p < 0.0001), slow postoperative mobility (p < 0.001),
and revision total hip arthroplasty (p = 0.0043) to be the only significant
risk factors for acute colonic pseudo-obstruction in their group of thirty-one
patients. Our inclusion of a control group allowed us to identify significant
risk factors associated with acute colonic pseudo-obstruction in patients who
had a total hip or knee arthroplasty; many of these risk factors have been
associated with acute colonic pseudo-obstruction in previous reports on
surgery in the
literature8,9,12,13,18-20
(see Appendix). In our series, advanced age was a significant risk factor for
the development of acute colonic pseudo-obstruction (p < 0.001), which is
consistent with the findings previously reported in the
literature8,11.
Patients of advanced age may have overall poorer health, may use more
medications that decrease gut motility, and may have a more frequent history
of gastrointestinal surgery, all of which increase the risk for the
development of acute colonic pseudo-obstruction and likely confound increased
age alone as a risk factor. Male gender, also a known risk factor for the
development of Ogilvie syndrome, was identified in the present series (odds
ratio = 3.75, p < 0.01). Although the reason for this finding is unknown,
most reports on the syndrome have described a 2:1 male-to-female ratio of
patients with acute colonic pseudo-obstruction, which is in agreement with our
findings8,17,22,23.
Seventeen of our eighteen patients (p < 0.0001) had at least one risk
factor for the development of acute colonic pseudo-obstruction besides
advanced age and male gender, with previous abdominal surgery being relatively
common (seven patients). Patients in the Ogilvie syndrome group were three
times as likely as the subjects in the control group to be on chronic
medications that are known to decrease gastrointestinal tract motility; the
difference was significant (p < 0.001). This finding warrants careful
evaluation of preoperative medications by the surgeon to help to identify
patients who are more likely to have Ogilvie syndrome develop in the
postoperative period.
In the current series, abdominal distension occurred in every patient with
acute colonic pseudo-obstruction. Abdominal discomfort, nausea, and vomiting
were all very common symptoms as well. Clarke et al. also reported abdominal
distension as the predominant symptom associated with acute colonic
pseudo-obstruction, with the symptom occurring in twenty-seven (90%) of the
thirty patients with acute colonic
pseudoobstruction8.
Interestingly, there was a return of flatus and bowel movements prior to the
onset of symptoms of acute colonic pseudo-obstruction in the majority (89% and
78%, respectively) of patients in our study, which mirrors the findings in the
study by Clarke et al. (76% and 81%, respectively). The flatus and bowel
movements are likely from the contents of the colon distal to the developing
pseudo-obstruction. Therefore, it is apparent that, although the return of
flatus and/or bowel movements heralds the return of some gastrointestinal
function after surgery, one should not assume that normal bowel function has
been restored.
In the present study, we found a trend between the use of
patient-controlled analgesia and the onset of symptoms (p = 0.067). Symptoms
developed twice as quickly in the patients with acute colonic
pseudo-obstruction who received patient-controlled analgesia (median, 2.2
days) as they did in the patients who did not (median, 4.6 days). Studies in
the literature are divided with regard to an association between the use of
patient-controlled analgesia and the development of acute colonic
pseudo-obstruction, with some
supporting9,19
and some refuting12
such a relationship. Unfortunately, we were not able to determine with
reliability the total amount of intravenous narcotic administered, whether
with patient-controlled analgesia or otherwise, from many of the charts.
The results of the present study showed that decompressive colonoscopy
should be strongly considered for any patient with acute colonic
pseudo-obstruction who does not respond to the initial nonoperative treatment
(Fig. 3). The patients who
underwent decompressive colonoscopy had a significantly shorter length of
hospitalization than those who did not have this procedure (p = 0.019). Few
studies have quantified the benefit of this procedure beyond the simple
acknowledgment of the resolution of symptoms.
Colonic perforation, the most feared sequela of acute colonic
pseudo-obstruction, did not occur in any of our patients. Traditionally, an
absolute cecal diameter of 12 cm has been associated with an increased risk of
perforation2,3,6,18,22.
Other authors have suggested that the risk of perforation is related more to
the rate and duration of cecal
dilatation17. In
previous studies of acute colonic pseudo-obstruction after arthroplasty,
complications have included death. These deaths were not the result of colonic
perforation but, rather, were due to other medical complications that may or
may not have been influenced by the occurrence of acute colonic
pseudo-obstruction8,9.
Most patients in whom acute colonic pseudo-obstruction develops have risk
factors that can be identified preoperatively. Knowledge of these risk factors
enables the surgeon to anticipate which patients may be at higher risk for the
development of acute colonic pseudo-obstruction. The resumption of flatus and
bowel movements postoperatively does not necessarily guarantee the return of
normal bowel function, and, therefore, any patient who complains of increased
abdominal distension should have a plain radiograph of the abdomen. Once the
diagnosis of acute colonic pseudo-obstruction is made, the nonoperative
measures reviewed in the present report should be the initial treatment. If
the colonic diameter nears 12 cm, or if symptoms are refractory to
nonoperative treatment, decompressive colonoscopy should be considered.
Tables listing risk factors for Ogilvie syndrome and drugs associated with
decreased gut motility are available with the electronic versions of this
article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?