Background: The measurement of fluid balance (FB) for hospitalized patients has been a standard of nursing practice for many years, but its utility and accuracy is questioned in the literature. There is no gold standard benchmark for accuracy of FB measurement. From a clinician’s standpoint, FB data are essential in guiding treatment decisions.  The inaccuracy of FB measurements directly impacts patient safety, patient flow, clinical reasoning and nursing/provider job satisfaction.

Purpose: Our project goal was to develop a process to improve the accuracy and documentation of FB monitoring and daily weights (DW) on a general medicine inpatient ward.  We used the concordance of Ins and Outs (IO) data and daily weight change (DWC) as a surrogate measure for accuracy of FB measurement.  A manual review of charts looking at data trends over the course of an admission revealed a concordance rate of 15%. FB and DW documentation were inadequately charted, 69% and 72% of admission days, respectively.  40% of charts did not have enough data to determine overall accuracy.

Description: A multidisciplinary team analyzed over 25 sources of error for the FB measurement process alone (See Figures). The electronic tracking tool relied on multiple different paper recording sites with subsequent entry into a cumbersome software program. Nurses then exported the data into the EMR. Gap analysis revealed inconsistent use of the tracking tool, loss of FB data during off-ward procedures, inaccurate DW measurement, inconsistent water pitcher measurements, and unclear delineation of nursing responsibility as the highest yield, most wasteful processes.  We developed a nursing driven protocol delineating responsibility for tracking IO data in a streamlined, two-step process.  A single paper bedside flowsheet is used to track FB data. An electronic nursing shift IO note is then generated at the end of each shift to summarize data in the EMR.  The DW process was standardized.  Patients were educated on using an improved graduated pitcher, thereby becoming active stakeholders in the process.  A nursing handoff process for shift change and off-ward procedures was implemented to capture more FB data.

Conclusions: Our pilot data showed an improved FB and DWC concordance rate from 15% to 41%.  Documentation rates for FB and DW improved to 76% and 84% of admission days, respectively.  Charts without enough data to determine accuracy improved from 40% to 21%.  Pre- and post-implementation surveys of nurses and providers revealed improvement in perceived accuracy, ease of making clinical decisions, ease of measurement, reporting, and locating data. Our system redesign demonstrated that measurement of FB is extremely complex and changes in process are hindered by a pervasive attitude that this is a simple task. Through interdisciplinary collaboration, we attained significant improvement in accuracy, nursing and provider satisfaction, and ease of using the process.