216: Relationship between ICU Admissions and Severe Maternal Morbidity


      Maternal ICU admission as a marker of severe maternal morbidity (SMM) has been identified by the Joint Commission and others as an indicator for multidisciplinary review. We sought more data on the incidence and characteristics of maternal ICU admission and its relationship to SMM in a large cohort of California maternal hospital admissions.

      Study Design

      As part of a larger study examining SMM, ICD9codes, complete blood bank records, prolonged hospital length of stay and maternal ICU data (using internal hospital sources) were obtained for all women giving birth from 16 California representative hospitals from July 1, 2012 - June 30, 2013. Medical charts were reviewed and true SMM diagnosed based on expert panel agreement using a clinical Gold Standard.


      Among 67,468 deliveries, there were 176 maternal ICU admissions with an overall ICU admission rate of 2.6/1000 deliveries(range:0.05-0.40). Higher maternal level of care hospitals had higher rates of ICU admission but each level showed great variation and overlap with other levels (TABLE). The majority of the ICU admissions (150/176 or 85%) had SMM by gold standard criteria. A greater proportion of ICU admissions in lower level hospitals were not for true SMM. Over one third 176/491(36%) of true positive SMM cases were admitted to the ICU. Of the 150 true positive cases amongst ICU admissions, 44% of severe morbidity was obstetric hemorrhage (25% of these were placenta complications, the remainder were atony or lacerations), followed by hypertensive disorders (28%). 58%(102/176) of ICU admissions received at least 1 unit of transfusion and 79% (81/102) of which were massive transfusions (≥ 4 units of blood. Nearly half (47%) of ICU admissions were ≤ 24 hours duration, 18% were 2 days, and 35% were ≥ 3 days.


      Although ICU admission was a very good predictor of true SMM, its accuracy varied substantially among hospitals. Higher maternal level of care hospitals had higher levels of both maternal ICU admissions and admissions with true SMM. Hemorrhage and hypertensive diseases were responsible for the largest proportion of ICU admission in this cohort.
      Figure thumbnail fx1