ABSTRACT
OBJECTIVE: To examine the ability of three different proteinuria assessment methods (urinary dipstick, spot urine protein:creatinine ratio [Pr/Cr], and 24-hour urine collection) to predict adverse pregnancy outcomes. METHODS: We performed a prospective multicentre cohort study, PIERS (Preeclampsia Integrated Estimate of RiSk), in seven academic tertiary maternity centres practising expectant management of preeclampsia remote from term in Canada, New Zealand, and Australia. Eligible women were those admitted with preeclampsia who had at least one antenatal proteinuria assessment by urinary dipstick, spot urine Pr/Cr ratio, and/or 24-hour urine collection. Proteinuria assessment was done either visually at the bedside (by dipstick) or by hospital clinical laboratories for spot urine Pr/Cr and 24-hour urine collection. We calculated receiver operating characteristic area under the curve (95% CI) for each proteinuria method and each of the combined adverse maternal outcomes (within 48 hours) or adverse perinatal outcomes (at any time). Models with AUC ≥ 0.70 were considered of interest. Analyses were run for all women who had each type of proteinuria assessment and for a cohort of women ("ALL measures") who had all three proteinuria assessments. RESULTS: More women were proteinuric by urinary dipstick (≥ 2+, 61.4%) than by spot urine Pr/Cr (≥ 30 g/mol, 50.4%) or 24-hour urine collection (≥ 0.3g/d, 34.7%). Each proteinuria measure evaluated had some discriminative power, and dipstick proteinuria (categorical) performed as well as other methods. No single method was predictive of adverse perinatal outcome. CONCLUSION: The measured amount of proteinuria should not be used in isolation for decision-making in women with preeclampsia. Dipstick proteinuria performs as well as other methods of assessing proteinuria for prediction of adverse events.