Laboratory abnormalities are shown in Table 2. The incidence of atypical lymphocytosis was significantly increased in VE 4- patients, compared with VE — patients (55 percent [16/29] vs 30 percent [26/86], respectively; p<0.05), especially when considering a threshold above 200/cu mm (31 percent [9/29] vs 7 percent [6/86], respectively; p<0.01). Frank mono- nucleosis-like reaction was observed in four of six patients with isolated viruria. Elevations in AST greater than one times normal and in LDH greater than two times normal were significantly more frequent in VE + patients; the same trend was observed with jaundice, although it was not significant. Acute renal failure occurred in 16 (55 percent) of 29 VE + patients and in 27 (31 percent) of the 86 VE — patients (p<0.05).
Two protocols for treatment are currently used for acute mediastinitis in our unit, depending on the size and depth of d6bridement. The first technique consists of twice-daily dressings of the open wound, and the second is closed irrigation-lavage. In the group of patients treated with the latter method, evaluation of the local infection can be easily and objectively assessed, because failure of this method necessitates reoperation because of persistent signs of local or general sepsis. Of the 57 patients treated with closed irrigation-lavage in this study, 24 were definitively cured, and 33 required reoperation. Only three (13 percent) of 24 cured patients were viral shedders, compared to 12 (36 percent) of 33 patients who required reoperation (p<0.05). Thus, failure of local treatment with a need for reoperation occurred in 12 (80 percent) of 15 patients with virologically proven CMV infection.
The overall mortality during the study period was 41 percent (47/115): 37 percent (31/86) for VE- patients, and 55 percent (16/29) for VE+ patients (NS). When considering late mortality (ie, after the 15th day of hospitalization), a significant difference was observed in regard to the virologic status: 16 (55 percent) of 29 VE + patients subsequently died vs 18 (25 percent) of 73 VE – patients (p<0.01). The mean duration of hospitalization among the 68 survivors was 20 days longer in VE+ patients (69 ±36 days), compared to VE- patients (48±27 days) (p<0.05). In addition, 62 percent (8) of the 13 VE+ survivors required more than 60 days of hospitalization vs 27 percent (15) of the 55 VE – survivors (p<0.05). A similar trend concerning the late mortality and the duration of hospitalization among survivors was noted when taking into account the modalities of the initial local treatment (open dressing or closed irrigation- lavage). Prior serologic status did not modify the morbidity observed in VE + patients.