A year of University of Pennsylvania School of Nursing research points to a consistent finding: the staffing numbers hospitals track most closely to predict patient safety are less reliable than nurses’ own read.
Here is what Penn Nursing’s recent work has found:
1. A nurse’s read on staffing can beat the spreadsheet.
In a study of 1,269 units across 217 hospitals, Penn’s Eileen Lake, PhD, RN, found that on med-surg and stepdown units, nurses’ own assessments of whether staffing was adequate predicted patient falls more accurately than objective metrics such as registered nurse hours per patient day.
“Nurses are uniquely positioned to judge staffing adequacy because they see the real-time complexity of patient care that administrative headcounts often overlook,” Dr. Lake said, arguing for direct dialogue with bedside nurses over reliance on quantitative reports alone.
2. When staffing is measured in patients, the stakes are measured in lives.
A study published in April in Medical Care analyzed nearly 550,000 medical-surgical patients across 132 Pennsylvania hospitals and found that each additional patient added to a nurse’s assignment carried 8% higher odds of death within 30 days, along with 4% higher odds of readmission and a 2% longer stay. The same added patient raised nurses’ odds of burnout 33%, job dissatisfaction 43% and intent to leave 27%.
Pennsylvania med-surg nurses average 5.9 patients each; moving to a 4-to-1 ratio could save an estimated $66 million in turnover costs and $239 million in shorter stays annually, the researchers projected.
3. Nurses leave over the same gap — and would return if it closed.
A February study in JAMA Network Open surveyed more than 4,000 RNs who left bedside roles between 2019 and 2023 and found most would consider returning. The top factors that would bring them back were adequate staffing (65%), flexible scheduling (59%) and better wages and benefits (59%).
“High nurse turnover is a solvable crisis, because the reasons nurses leave are the same reasons they would return, if addressed,” said lead author Karen Lasater, PhD, RN.
4. In psychiatric units, the same lever holds.
A Penn study of 740 inpatient psychiatric nurses found 38% reported high burnout and nearly 1 in 4 intended to leave within a year. Each one-point improvement in a unit’s work-environment score cut the odds of high burnout by about 70% and intent to leave by 53%. Staffing adequacy was the only work-environment domain significantly tied to all three outcomes measured — burnout, intent to leave and dissatisfaction — and it was also the lowest rated.
5. Technology is not a shortcut around staffing.
Two studies caution leaders against expecting new tools to paper over thin coverage. Bedside nurses gave virtual nursing programs mixed reviews: 57% said the model did not reduce their workload and 10% said it increased it, and researchers concluded such programs are unlikely to help “without sufficient bedside staffing and clearly defined roles.”
A separate Penn analysis warned that AI may add hidden costs and workflow burden rather than remove it, flagging “algorithmic drift” and noting some tools “may even increase clinicians’ cognitive burden or documentation workload.” In both cases, the researchers’ recommended safeguard was the same: bring nurses into the design and rollout.
Penn’s researchers increasingly frame the bedside nurse’s judgment about staffing not as a soft data point to be reconciled against the headcount, but as one of the most accurate early signals hospitals have for patient safety and workforce stability.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
