patients (10,11). Nevertheless, the efficacy of this technique is real: Cinat et al. found this
method to be 90% successful in postoperative abscess. Abscesses involving the appendix, liver
or biliary tract, and colon or rectum were also found to be particularly responsive at rates of
95%, 85%, and 78%, respectively, although pancreatic abscesses and those involving yeast were
correlated with poor outcomes by this treatment method (10). Khurrum Baig et al. echoed the
success of percutaneous drainage in treating abscesses secondary to colorectal surgery, but
questioned the applicability of these findings to patients with other than well-defined intra-
abdominal abscesses (11).
Other considerations include planned relaparotomy and open management. Data is far
from optimal, as these critically ill patients cannot ethically be randomized to different
treatment groups. However, it would appear at this time that these strategies still are
associated with a high mortality of around 42% (12,13). A study by Schein found a particularly
high mortality of 55% in the specific subgroup of diffuse postoperative peritonitis treated by
planned relaparotomy, with or without open management. Furthermore, Schein went on to
state that open management was associated with over twice the mortality of closed: 58% versus
24% (14). Although necessary flaws in study design make it difficult to say whether these
approaches offer an advantage over the more traditional ones, it is nevertheless clear that they
are far from ideal.
The hurdles in addressing the challenge of tertiary peritonitis have led to exploration of
potential future therapies. Some are in keeping with traditional surgical/mechanical means:
Case studies have reported success of laparoscopy, even in the face of diffuse peritonitis and
multiple abscesses (15). Other concepts favor a medicine-based approach, rooted in emerging
ideas on the disease’s basic pathology. As it is believed that bacteria migrate out of the
intestinal tract secondary to mucosal ischemia and permeability, strategies that support the
mucosa, such as early postoperative enteral feeding or selective elimination of endogenous
pathogenic bacteria, have each been tried with mixed results. Likewise, it has been argued that
the progression from secondary to tertiary peritonitis represents a crippling of the body’s
immune system; in support of this belief, granulocyte colony–stimulating factor and
interferon-c have each produced limited success in small patient groups, and successfully
treated individuals all demonstrated some recovery of immune cell functioning. Another
postulate is that a relative lack of corticosteroid exists to fulfill the demands of extreme stress,
and it has been suggested that supplying some patients with stress doses of hydrocortisone can
improve the vascular effects in early sepsis. Modulation of the inflammatory cascade with
activated protein C continues to be investigated, including the associated risk of bleeding.
Finally, some researchers have examined the possibility that alleviating the hyper-catabolic
state of patients with tertiary peritonitis might decrease mortality. Growth hormone and
insulin-like growth factor-1 have both been tried with intermittent positive and negative
outcomes (9).
NEW-ONSET PERITONITIS
Antibiotic-AssociatedClostridium difficileDiarrhea in the ICU Patient
Epidemiology, Pathogenesis, and Risk Factors
The anaerobeC. difficilecauses twice as many cases of diarrhea as all other bacterial and
protozoal causes combined. In hospitalized patients,C. difficileis responsible for 30% of
diarrhea cases, and in hospitalized patients receiving antibiotic therapy—as is the case for
many postsurgical patients—this number rises to an impressive 50% to 70%.C. difficile–
associated diarrhea (CDAD) is theorized to arise in patients colonized by the pathogen when
protective normal gut flora is simultaneously suppressed by broad-coverage antibiotic
exposure. Although clindamycin, ampicillin, and the third-generation cephalosporins such as
ceftazidime, ceftriaxone, and cefotaxime are the most commonly associated antimicrobials, the
newer, broader spectrum quinolones, such as gatifloxacin and moxifloxacin, can also increase
risk, and in fact any antibiotic, including, surprisingly, metronidazole and vancomycin, may
rarely predispose patients to the disease. Other risk factors for CDAD include age,>60 years,
the winter season, antineoplastic agents (especially methotrexate), recent gastrointestinal
surgery, enemas, stool softeners, postpyloric enteric tube feedings (e.g., J-tubes), and even use
of proton-pump inhibitors in hospitalized patients (16,17).
262 Wilson