Methods: Patients who sustained a fall between October 1, 2014 and September 31, 2015 were identified from the institutional incident report database. The electronic health records of 50 of those patients were randomly selected to be reviewed retrospectively. Data collected included patient age, gender, the date and time of all Morse Fall scores, including the total score and the score for each individual item, time and date of any falls, and dates of any transfers between units. Any score which was a change, either increasing or decreasing, was compared to documentation in the progress notes. In the event of a discrepancy, the progress note data was identified as valid and a corrected score was calculated. For example, if the MFS indicated no IV, but the progress note documented an IV dressing or change, the appropriate points were added to the score and error was flagged. All raw data, including both documented and corrected scores, when applicable, were entered into an Excel spreadsheet. All data was rechecked for accuracy, then de-identified. The spread sheet was password protected and stored behind the VA firewall. Scores were dichotomized into those below 45 and those above 45, the high risk cutoff at our institution. Each patient’s scores were reviewed for any significant change, defined as a raw score change sufficient to move that patient into the other group.
Results: After correcting for calculation errors, 70% of patients had no change in their MFS, having scored in the high risk group on admission and remaining there throughout their hospital stay. Four scores (8%) changed to high risk due to a fall. Two scores (4%) increased due to a change in mental status. Six patients (12%) had changes in their scores due to changes in the nursing assessment of gait, or the presence of an IV. Generally, these changes resulted in a temporary decrease, briefly bringing them out of the high risk group. For 50 patients, there were 890 MSF assessment, estimated to represent 75 hours of nursing time. There was a discrepancy in 361 (41%) of the assessments. 5.4% of the discrepancies were due to errors in addition. Other errors included omitting history of falls, comorbidities, presence of an IV, ambulatory aids, and mental status issues, such as confusion. Gait assessment fluctuations were commonly due to assigning 0 points for bedrest during the night shift.
Conclusion: There is little change in the corrected MFS score of adult patients during their hospitalization. Frequent repetition of the MSF assessment consumes nursing time, a valuable and limited resource. Performance of the MSF assessment during the night shift is particularly problematic . Our data suggests that assessment on admission, transfer, after a fall or any change in mental status would be a more efficient approach to identifying patients at risk and that, once a patient is identified as at high risk, they should remain there and not require any further MFS assessments.
See more of: Evidence-Based Practice Sessions: Oral Paper & Posters