NR 558 Week 4 Decision Support Rule Critique Example

Clinical Decision Support Critique: The High Fall Risk Rule That Fires at a Morse Score of 45, Its Logic as Configured, the 1989 Source It Was Built From, and Where It Reaches Wider or Narrower Than the Evidence

Student Name

College of Nursing, Chamberlain University

NR 558: Informatics for Professional Nursing Practice

Instructor Name

Month Day, Year

Clinical Decision Support Critique: The High Fall Risk Rule That Fires at a Morse Score of 45, Its Logic as Configured, the 1989 Source It Was Built From, and Where It Reaches Wider or Narrower Than the Evidence

The Rule as Configured

Our electronic health record contains a clinical decision support rule, named "High Fall Risk Protocol," that works as follows. Condition: when the Morse Fall Scale total filed in the shift assessment is 45 or higher. Elements read: the six Morse item scores and the calculated total, and the patient's location. Group reached: all adult inpatients on medical, surgical and telemetry units; not intensive care, behavioral health, obstetrics or pediatrics. Action when met: a yellow banner appears on the patient's chart header reading "High Fall Risk"; an order set is suggested to the nurse containing a bed alarm, a yellow wristband and nonskid socks, which the nurse can accept with one click; and a task is added to the nurse's worklist to complete fall prevention education. Action when the score drops below 45: the banner disappears at the next filed assessment, but the orders remain until discontinued.

Written this way, a stranger could restate the logic without seeing the build: a single number crossing a single threshold triggers a banner, a suggested order set and a task.

The Source, With a Year on It

The rule was built from the Morse Fall Scale, first published in 1989 (Morse et al., 1989). The scale assigns points for six factors: history of falling, secondary diagnosis, use of an ambulatory aid, intravenous therapy or heparin lock, impaired gait and mental status in the sense of overestimating one's own ability. Totals can range from 0 to 125. The threshold of 45 used by our rule is one commonly used cutoff for high risk, but the original work and later users have treated cutoffs as something to be calibrated for each setting rather than a universal value. The informatics team's build documentation, dated six years ago, cites the scale and a nursing policy but no local calibration.

Since 1989, evidence on hospital falls has moved on. A systematic review found that fall risk assessment tools, including Morse, had modest and variable predictive accuracy in hospital inpatients, and that identified risk factors, rather than a score alone, may be more useful for targeting prevention (Oliver et al., 2004). A later review of fall prevention in hospitals emphasized multifactorial interventions tailored to the individual rather than generic bundles triggered by a score (Oliver et al., 2010).

Narrower Than the Source

The rule is narrower than the evidence in two ways. First, it reads only the total, discarding which items produced it. A patient who scores 45 because of intravenous access and a secondary diagnosis gets the same response as a patient who scores 45 because of confusion and an unsteady gait, although their risks call for different interventions. The scale's items carry clinical information that the rule throws away. Second, it excludes observation patients, although they lie in the same beds with the same hazards, because they are not formally inpatients.

Wider Than the Source

The rule is wider than the evidence in one important way. It triggers the same order set for everyone above the threshold, regardless of the reason. Bed alarms, for instance, have limited evidence of benefit and can contribute to alarm fatigue and immobility, and the reviews above favor individualized interventions. Applying a uniform bundle to every high score reaches further than the evidence supports.

How the Rule Interacts With Copy-Forward

The rule's behavior depends on the accuracy of its input. Since each shift inherits the prior shift's Morse answers unless someone edits them, the rule is often reading an old assessment. A patient who becomes confused overnight and is started on intravenous fluids may still show a score below 45 the next morning if the nurse files without updating, and the rule will not fire. Conversely, a patient whose intravenous line was removed and who is walking steadily may keep a high score, a banner and a bed alarm for days. The rule is only as current as the least current item it reads.

Alert Burden

The banner fires for about 40% of patients on our unit at any time, and the suggested order set appears at every filed assessment while the score remains 45 or above. Nurses report accepting the order set by habit. Research on decision support has linked repeated alerts are associated with declining responsiveness (Ancker et al., 2017), and a banner seen on almost half the charts on the unit conveys little. A rule designed to identify the highest-risk patients has become a label for a large share of them.

Who Owns the Rule

The rule was built by the clinical informatics team at the request of the fall prevention committee six years ago and approved by the nursing informatics council. No one is currently assigned to review it. Its documentation lists the Morse scale and the hospital's fall policy as sources but records no validation against local falls. A rule without an owner or review date keeps running whatever changes around it, including changes in the evidence, the patient population and the documentation that feeds it. Assigning an owner and an annual review would be the simplest improvement of all.

What the Critique Recommends

Four changes follow. First, fix the input: make the items most likely to change arrive blank each shift, so the rule reads a current assessment. Second, display which items produced the score in the banner, for example "High fall risk: confusion, unsteady gait," so the response can match the reason. Third, replace the uniform order set with a short menu of interventions linked to the items, such as toileting schedules for urgency or supervised mobility for gait, consistent with evidence for individualized prevention. Fourth, calibrate the threshold locally by comparing scores with falls on our units over a year, and review the rule on a set schedule, since the source it was built from is more than three decades old.

A rule this visible, on nearly half the unit's charts, deserves at least as much attention as the policy it was built to support.

The informatics team has offered to pull a year of scores and falls for the calibration.

Results will go to the council.

References

Ancker, J. S., Edwards, A., Nosal, S., Hauser, D., Mauer, E., & Kaushal, R. (2017). Effects of workload, work complexity, and repeated alerts on alert fatigue in a clinical decision support system. BMC Medical Informatics and Decision Making, 17, Article 36. https://doi.org/10.1186/s12911-017-0430-8

Morse, J. M., Morse, R. M., & Tylko, S. J. (1989). Development of a scale to identify the fall-prone patient. Canadian Journal on Aging, 8(4), 366-377. https://doi.org/10.1017/S0714980800008576

Oliver, D., Daly, F., Martin, F. C., & McMurdo, M. E. T. (2004). Risk factors and risk assessment tools for falls in hospital in-patients: A systematic review. Age and Ageing, 33(2), 122-130. https://doi.org/10.1093/ageing/afh017

Oliver, D., Healey, F., & Haines, T. P. (2010). Preventing falls and fall-related injuries in hospitals. Clinics in Geriatric Medicine, 26(4), 645-692. https://doi.org/10.1016/j.cger.2010.06.005