D ATA D R I V E N . P O L I C Y F O C U S E D LDI ResearchBRIEF Research to Improve the Nation’s Health System 2020 . No. 1 CHRONIC HOSPITAL NURSE E D BY A P E UNDERSTAFFING MEETS COVID-19 RT N PO N COVID-19 SUP An Observational Study LDI RAPID RESPONSE GRANT Karen B. Lasater, Linda H. Aiken, Douglas M. Sloane, Rachel French, Brendan Martin, Kyrani Reneau, Maryann Alexander, Matthew D. McHugh BMJ Quality & Safety — published online August 18, 2020 KEYFINDINGS A study of hospitals in New York and Illinois at the start of the COVID-19 pandemic found that most did not meet benchmark patient-to-nurse staffing ratios for medical-surgical or intensive care units. New York City hospitals had especially low staffing ratios. Understaffed hospitals were associated with less job satisfaction among nurses, unfavorable grades for patient safety and quality of care, and hesitance by nurses and patients to recommend their hospitals. THE QUESTION THE FINDINGS Better hospital nurse staffing is associated with lower patient mortality, The study included 4,298 registered nurses (RNs) working in medical- fewer complications, high patient satisfaction scores, shorter lengths of surgical units at 254 hospitals and 2,182 RNs in ICUs at 179 hospitals. stay, and fewer readmissions. However, staffing levels vary widely across Patient-to-nurse ratios in medical-surgical units ranged from 3.3 to 9.7. hospitals, and only one state (California) has comprehensive legislation Understaffing was even more acute in the New York City area, with delimiting how many patients a nurse can care for at one time. In the 20 average patient-to-nurse ratios of 6.5 in medical-surgical units (Figure 1). years since California’s mandate, similar proposals, such as a 2018 initiative Staffing ratios in ICUs also varied, ranging from 1.5 to 4, with New York in Massachusetts, have failed. City hospitals worse than hospitals elsewhere. A principal policy barrier is a lack of timely data regarding local staffing Figure 1. conditions. Staffing requirements come at significant cost to hospitals, Hospital variation in medical-surgical staffing and policymakers often have limited insight into the consequences of nurse staffing levels, which complicates weighing policy tradeoffs. In this 10 study, the authors surveyed nurses in Illinois and New York—two states 9 considering staffing legislation—to determine the variation of nurse 8 staffing levels in medical-surgical and intensive care units (ICUs) and Patients/Nurse Ratio 7 estimate the association with quality of care, patient satisfaction, and 6 nurse job outcomes. 5 4 3 2 1 0 Hospitals arrayed by Staffing| Mean: 5.56; Range: 3.28 - 9.67 NYC, Nassau and Westchester Counties Illinois & Rest of New York COLONIAL PENN CENTER | 3641 LOCUST WALK | PHILADELPHIA, PA 19104-6218 | LDI.UPENN.EDU | P: 215-898-5611 | F: 215-898-0229 | @PENNLDI ResearchBRIEF LDI Figure 2. Proposed legislation in both states would Odds of poor outcomes reported by medical-surgical nurses for each additional patient per nurse mandate a maximum of 4 adult medical-surgical patients and 2 ICU patients per nurse per shift. Safety and quality outcomes This study reveals a wide variation in current average staffing levels, and nurses working in Not "Definitely Recommend Hospital" hospitals that regularly exceed these thresholds. Poor/Fair Quality of Care The findings comport with other reporting Unfavorable Patient Safety Grade related to COVID-19: variation in access to Unfavorable Infection Prevention Grade resources, such as RNs, can lead to unequal outcomes for patients. Missed Patient Surveillance Missed Administering Medications on Time The COVID-19 pandemic has seriously affected hospital budgets, raising concerns about the Important Patient Care Information is Lost During Handoffs cost of staffing mandates. This timely analysis Nurse outcomes shows the other side of the tradeoff. In two Job Dissatisfaction states considering legislation to set minimum Intent to Leave Their Job requirements, the variation in hospital nurse High Burnout staffing adversely affects quality of care and patient safety, as well as outcomes for RNs. 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Further research is needed to evaluate other Odds Ratio relevant policy considerations, such as the scale of the cost required to comply with new regulations. Overall, about half of nurses across all settings reported burnout and THE STUDY unfavorable patient safety grades, about 70% would not “definitely Survey data were collected from December 2019 through February recommend” their hospital, and two thirds reported frequent work 2020 from direct-care RNs in New York and Illinois hospitals. The survey interruptions and delays due to insufficient staff. About a third of patients questions used previously validated measures of personal characteristics gave poor satisfaction ratings and would not definitely recommend their (e.g., years of experience), job outcomes (e.g., burnout), and quality of hospital. Across most measures, surveys revealed worse outcomes in New care and patient safety at their institution. These surveys were aggregated York City hospitals. at the hospital level to create hospital-specific measures of each variable, These outcomes were highly correlated with staffing levels (Figure 2). including the average number of nurses per patient per shift. After controlling for hospital factors (e.g., size and teaching status), nurse Staffing data were linked to patient satisfaction data from the Hospital characteristics (e.g., years of experience), and geography, each additional Consumer Assessment of Healthcare Providers and Systems (HCAHPS) patient per nurse was associated with significant increases in the odds and hospital characteristics gathered from the American Hospital of nurses reporting poor outcomes. Furthermore, patient-reported Association (AHA) Annual Survey. Hospital and personal characteristics satisfaction measures were also associated with staffing and validated were used as controls when modeling the effect of staffing levels on nurse-reported measures. patient and nurse-reported outcomes. THE IMPLICATIONS Lasater KB, Aiken LH, Sloane DM, French R, Martin B, Reneau K, Both New York and Illinois had severe COVID-19 outbreaks at the Alexander M, McHugh MD. (2020). Chronic Hospital Nurse Understaffing start of the pandemic. This study demonstrates that even before the Meets COVID-19: An Observational Study. BMJ Q&S. doi: 10.1136/ pandemic strained health care systems, hospital nurses were already bmjqs-2020-011512 working in understaffed conditions, leading to frequent process failures that jeopardized patient safety, and contributed to emotional distress and This study was supported by a Penn LDI COVID-19 Rapid Response Grant. high turnover among nurses. Patient reviews of hospitals corroborated This grant supports LDI Senior and Associate Fellow projects that address these findings. Both patient and nurse assessments of hospital quality were urgent questions related to population health, health care delivery, or health strongly associated with nurse staffing levels in these states. policy during the pandemic. THE RESEARCHER KAREN LASATER, PhD, RN Karen Lasater is an LDI Senior Fellow and Assistant Professor at Penn's School of Nursing in the Center for Health Outcomes and Policy Research. Her program of research utilizes a health services research framework and quantitative methods to study the impact of nursing on patient outcomes.