Discussion

Prior civilian trauma population studies demonstrated incidences of CDAD similar to nontrauma critically ill patients (1%–3% and 4%–5%, respectively) despite a lack of traditional risk factors.^{Reference Vanzant, Ozrazgat-Baslanti and Liu3,Reference Lumpkins, Bochicchio and Joshi4} Despite widespread antimicrobial use in our military trauma population, CDAD rates were low (0.86%; 2.76 per 10,000 OBD). A study of cumulative antibiotic exposure in hospitalized patients showed a median of 14 antibiotic days in CDAD patients,^{Reference Stevens, Dumyati, Fine, Fisher and van Wijngaarden7} similar to the median of 13 antibiotic days identified in our population. The same study identified a median of 3 antibiotic class exposures in the CDAD group, while our CDAD patients received a median of 6 antibiotic class exposures prior to diagnosis. More than half of patients were exposed to ≥5 antibiotic classes prior to CDAD diagnosis, which is consistent with literature suggesting higher CDAD risk with increased number of antibiotic class exposures.^{Reference Stevens, Dumyati, Fine, Fisher and van Wijngaarden7} We noted extensive tetracycline exposure in our population, largely driven by malaria prophylaxis. Although tetracyclines have demonstrated lower risk for CDAD than other antibiotics, given the high degree of broad antimicrobial class exposure, current data are insufficient to determine whether tetracycline exposure impacted our population.^{Reference Tariq, Cho and Kapoor8}

Similar to civilian trauma cohorts,^{Reference Vanzant, Ozrazgat-Baslanti and Liu3,Reference Lumpkins, Bochicchio and Joshi4} our CDAD patients were severely injured and were primarily admitted to the ICU, were intubated prior to diagnosis, and were diagnosed >1 week into their hospitalizations. This high injury severity is characteristic of the overall TIDOS population: 64.5% with blast injuries; 38% with ISS ≥25; 52% admitted to the ICU; 33% mechanically ventilated.^{Reference Campbell, Li and Whitman9} Among our CDAD patients, 74% were diagnosed with ≥1 preceding infection, and 34% of the overall TIDOS population had ≥1 infection, primarily SSTIs.^{Reference Campbell, Li and Whitman9}

Patients were primarily treated with metronidazole due to contemporary literature and guidelines during the study period.^{Reference Napolitano and Edmiston2} Although CDAD was severe or fulminant in >50% of patients, there were no deaths. This finding differs dramatically from mortality rates in nontrauma patients (30%–80% in severe cases), but it is similar to that in civilian trauma studies and may result from our population being aged in their mid-20s and healthy prior to trauma.^{Reference Vanzant, Ozrazgat-Baslanti and Liu3,Reference Lumpkins, Bochicchio and Joshi4,Reference Karas, Enoch and Aliyu10} Although our study largely involved penetrating trauma (compared to blunt trauma in civilian studies),^{Reference Vanzant, Ozrazgat-Baslanti and Liu3,Reference Lumpkins, Bochicchio and Joshi4} we were unable to evaluate certain CDAD risk factors (eg, gastric acid suppression, enteral feeding, and intraabdominal surgery) and complications (eg, toxic megacolon or colonic perforation). Further investigation is needed to determine whether penetrating and blunt trauma populations have distinct risk factors for CDAD.