What happens when nursing teams are short-staffed?

Nurses are a hugely important staffing group within the NHS. There are more than 400,000 hospital nurses in England, the majority of whom work in teams that provide round-the-clock care for patients requiring hospital stays. They are responsible for monitoring patients, implementing and adjusting treatment plans, and coordinating overall care: all critically important tasks.

Over the coming years and decades, as the healthcare demands of an ageing population grow, the NHS is going to need to employ more nurses. But it also needs to make sure that it organises and deploys its existing nurses effectively, and to find ways to enhance their productivity. That means understanding how both the number and the mix of nursing staff affect the care provided to hospital patients.

In new research, recently published in the Journal of Health Economics, we examined the impact of nurse shortages on patient outcomes. We used data which combined detailed information on nursing staffing teams (e.g. who worked what shift, and who was missing) with the characteristics and clinical outcomes for the patients under their care (e.g. what they were diagnosed with, and what happened to them). This allowed us to explore what happens when a nursing team had fewer staff than they were supposed to, and compare the effect of an absence of a more qualified nurse with the effect of an absence of a less qualified nurse. All of the data came from a single large NHS hospital group in England, and was for the calendar year 2017.

The impacts of nurse shortages

There were four key findings from our analysis.

First, nursing teams that have insufficient numbers of qualified staff, relative to what was defined by managers as the safe level for the team, have poorer patient outcomes. For the average hospital ward in our sample, the absence of a registered nurse (one with university-degree-equivalent-level training) increases the odds of a patient death by approximately 10%. Patients treated by those teams are also more likely to be transferred to an intensive care unit (ICU) – one indication that their condition has deteriorated. In contrast, the absence of a nursing assistant, who lacks the formal training of a qualified nurse, has no impact on patient mortality or transfers to the ICU.

Second, experienced nursing teams achieve better outcomes. The absence of the most senior staff (as defined by their pay band) has a larger effect. And teams made up of staff with more experience working at the hospital in question achieve better outcomes, even after controlling for the size and seniority of the team. These effects are large: increasing the team’s average length of experience in the hospital by one year had the same impact as adding three-quarters of an extra nurse.

Third, the negative effects of nursing shortages are largest for patients diagnosed with sepsis. This is a condition that can develop suddenly and where early detection is of critical importance for survival, and where nurses play a central role in monitoring patients for the early signs. This suggests that reductions in the quality of monitoring and the timeliness of treatment, and potentially delays in escalation to a doctor, are key channels through which nurse staffing shortfalls affect patient mortality.

Fourth, and perhaps most surprisingly, the negative effects are concentrated among patients classified as being less severely ill (based on their diagnoses). This suggests that short-staffed teams focus their efforts on the patients who are most observably or obviously sick. Other patients, with less severe health conditions, receive less care and have worse outcomes as a result.

Conclusions

This research builds on the many previous papers showing the important role of nurses in the delivery of patient care. Our exceptionally granular and detailed data allow us to go beyond previous studies, and examine which types of nurses have the biggest impacts, and which patients are most affected.

One key conclusion relates to the mix of nursing staff. Increasing the supply of qualified nurses is expensive and takes time. Many healthcare systems, including the NHS, are seeking to make greater use of nursing assistants, nursing associates and other similar roles which can be filled more quickly and cheaply. Our results suggest that these are not straight substitutes for qualified nurses, and should not be treated as such.

Another conclusion relates to the returns to seniority and experience. Our findings suggest that problems arise when the nurse in charge of the ward is missing, and when the team has less experience working in that particular hospital. This latter finding points to the importance of improving staff retention, in order for nursing teams to build up such experience.

The fact that nurse shortages seem to particularly affect patients with sepsis will be important for clinical practitioners and workforce planners seeking to mitigate the negative impacts of staff shortages. Our results point to the possibility of considerable gains from protocols and training aimed at improving the ability of the broader nursing team to detect patient deterioration at an early stage.

Finally, this research is a useful case study for the value of linking together different NHS datasets. In this case, the link between detailed staffing roster data and electronic patient records allowed us to conduct high-frequency analysis that revealed new relationships and policy insights. Yet this sort of link is rare. We might all benefit from making such NHS data linkages routine.

This research was enabled by the Imperial Clinical Analytics Research and Evaluation (iCARE) environment and used the iCARE team and data resources. Data management was provided by the Big Data and Analytical Unit (BDAU) at the Institute of Global Health Innovation (IGHI).