Study Looks at Nurse Predictions and Patient Outcomes

Clinical nurses often presume what might happen with a patient: She’ll probably need assisted living upon discharge. I doubt he’ll head home today due to increasing ammonia levels. But what happens when nurses are asked to document such predictions about their patients’ outcomes? Mark McClelland, DNP, RN, CPHQ, a nurse scientist in Cleveland Clinic’s Office of Nursing Research and Innovation, conducted a phenomenological study to find out.

Mark McClelland, DNP, RN, CPHQ

“I believe that the nursing process mirrors the scientific process,” says Dr. McClelland. “I’ve got a strong interest in helping nurses develop a scientific approach to their care.” A very important part of a scientist’s work is making predictions, yet there is very little information in the literature about nurses making predictions about their patients’ outcomes.

Study participants follow a four-step process

Dr. McClelland’s team completed an initial study of 15 nurses on eight medical-surgical and step-down units in four hospitals. The purpose of the “Make Two Predictions” study was to explore the following questions:

• How do nurses perceive their ability to make clinically based predictions about a patient’s health outcomes?
• How do nurses making clinically based predictions change the care they provide?

Participants were asked to follow the Predict, Reconcile, Reflect and Learn (PRRL) Process and document it in a research log for two to three weeks. Developed by the research team, the PRRL Process is as follows:

• Predict – Make a prediction about two patients during each shift. Those predictions must be clinically based and identify a cause and effect. The results of the predictions also must be knowable within 48 hours.
• Reconcile – Review the patient’s records within 48 hours to determine if the prediction was accurate.
• Reflect – Think about the prediction and accuracy.
• Learn – Consider how and why you arrived at the answer you did. In particular, how did assumption-making influence your thinking?

During the data collection period, nurses made a variety of predictions – some accurate and others not. For instance, one prediction was that a patient would be able to control his blood sugar without insulin upon discharge. (That proved accurate.) Another prediction was that a patient would develop aspiration pneumonia. (That was not accurate.)

Results revealed five themes in nurse perceptions

Using a semi-structured interview guide, Dr. McClelland interviewed participants upon completion of data collection. Although the data is still being analyzed, five themes emerged.

1. Anticipation and proactivity – Using the PRRL Process caused nurses to think about future activities, particularly care coordination and discharge planning.
2. Empowerment – The process increased nurses’ confidence, which in turn improved their communication with the medical team and nonlicensed caregivers.
3. Expanded thinking – Nurses learned to think beyond the electronic medical record and medical plan of care.
4. Connecting the dots – Nurses took a more mindful view of patient care, reporting greater awareness and focus. “They did more thinking about care, instead of just doing it,” says Dr. McClelland.
5. Role definition – Participants reported becoming more aware of the role of nursing in creating patient outcomes.

Next steps for the study

The next steps are to further analyze and characterize the predictions made by the nurses in this study to create a prediction typology. Then Dr. McClelland would like to repeat the study engaging all the nurses on several units to learn how the social network affects the PRRL Process.

Ultimately, the PRRL Process could be a valuable educational tool for nurses. “Because we’ve gotten evidence that this changes thinking a bit, I believe the process may be used to assess a nurse’s competency and help teach younger nurses how to view patients differently,” says Dr. McClelland.

Photo credit Annie O’Neil.