A case review of BMT patients admitted to the ICU (a 16-bed medico-surgical ICU in a tertiary care teaching institution) over a 4-year period.
RESULTS: Between January 1994 and December 1998, 57 among 441 BMT patients (12.9%) were admitted to the ICU, mainly for respiratory distress (58%) and
hypotension (32%). Forty-two patients (73.7%) presented ARF as defined as a doubling of serum
creatinine. Compared to the 15 other patients, ARF patients had a higher APACHE II score (30 +/- 8 vs. 25 +/- 7, p < 0.05). For ARF vs. non-ARF patients, there was no difference in age (43.8 +/- 10.8 vs. 44.3 +/- 11.1 years), in requirement for
mechanical ventilation (76 vs. 73%) and vasopressors (69 vs. 60%), and in prevalence of
graft-versus-host disease (19 vs. 13%) or
neutropenia (69 vs. 67%), but the prevalence of
sepsis (83 vs. 60%) and
liver failure (69 vs. 40%) was higher. Maximum serum
bilirubin was markedly increased in ARF compared to non-ARF patients (p < 0.005). For both subgroups, no difference in the administration of potential nephrotoxic agents was identified. Usually, ARF was considered multifactorial by clinicians, with ATN being the most frequent diagnosis (55%). Maximum serum
creatinine reached a mean of 330 +/- 130 micromol/l. In 74% of cases, ARF occurred concomitantly or after admission to the ICU. Oligoanuria was present in 38%, whereas
polyuria was observed in 17%. Fourteen ARF patients (33%) required dialytic support. Mortality rates were significantly different in ARF vs. non-ARF patients (88 vs. 60%, p < 0.05). Predictive factors for the development of ARF were
liver failure (odds ratio (OR) 5.9), low
serum albumin (OR 1.2) and APACHE II score (OR 1.1), whereas variables predictive of mortality were
mechanical ventilation (OR 14.8), ARF (OR 5.8),
liver failure (OR 3.7), and APACHE II score (OR 1.2).
CONCLUSIONS: This study confirms that ARF in BMT patients admitted to the ICU is frequent, multifactorial, related to
liver failure, and that its development has a negative impact on outcome.