Introduction Current risk assessment guidelines for pre-eclampsia are unable to reliably predict adverse outcomes and thus inform management. This study investigated the association and predictive value of urinary albumin: creatinine ratio (ACR) in pregnancy for adverse outcomes.

Methods Demographic, clinical and biochemical data were collected, retrospectively, for 914 pregnant women who had an ACR test between December 2009 and February 2012. Fetal outcome and preterm delivery data were collected: other maternal outcomes were collected for a subset of 385 pregnancies. Odds ratios were used to investigate the association of ACR at two thresholds (30 and 60mg/mmol) with outcome. Sensitivity and specificity investigated the predictive ability. Proportions were compared using the Chi-square test.

Results In women with ACR > 30 mg/mmol the following outcomes were more frequent (P < 0.05): preterm delivery (38.2% vs. 10.5%), hypertension (95.7% vs. 69.4%), treatment with anti-hypertensives (71.3% vs. 30.2%), admission to maternal HDU (61.7% vs. 22.0%) and admission of the baby to neonatal unit (22.4% vs. 5.8%). There was a significant association between ACR > 30 mg/mmol and these outcomes (OR >1, 95% CI not crossing 1) but predictive ability was limited; specificity (73–96%) and sensitivity (31–66%). At a higher threshold, ACR >60mg/mmol did not demonstrate improved sensitivity (21–44%).

Conclusion Raised ACR appears to be associated with an increased likelihood of developing certain adverse outcomes. However the predictive ability is limited by poor sensitivity.