Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Study looks at nursing assistants' effect on patient outcomes

  • 1 Comment

“Patients are a fifth more likely to die on wards where nurses have been replaced by untrained staff, a major study has found,” the Daily Mail reports.

This latest research into 243 hospitals across Europe found those with more professional nurses, compared to nursing assistants, had lower death rates after surgery and were rated more highly by nurses and patients.

The numbers of professional nurses who’d had at least three years of training, as a proportion of all nursing staff, ranged from 82% in Germany to 57% in the UK. The researchers calculated that every increase of 10% in proportion of qualified nurses was linked to an 11% lower risk of death for patients after surgery.

However, analysis of the data is complicated by differences from country to country and interpreting it is more complex than media headlines suggest.

It is unclear whether training for nursing assistants was equivalent in all countries studied. In England, the Department of Health plans to introduce “nursing associates”, who would have 18 months’ training and work alongside professional nurses and existing, less well-trained health care assistants.

Additionally, the study does not prove that more qualified nurses are the reason for the differences in death rates and quality of care. The research is based on one “snapshot” of what was happening in hospitals at one point in time (2009 to 2010). Other factors, such as local doctor staffing levels, may also have an effect on outcomes.

Where did the story come from?

The study was carried out by researchers from the University of Pennsylvania School of Nursing, University of Southampton, Kings College London, University of Leuven in Belgium, Technische Universitat Berlin, Instituto de Salud Carlos III in Spain and Institute of Nursing Science in Basel.

It was funded by the European Union, National Institute of Health Research, National Institutes of Health and Spanish Ministry of Science and Technology.

The study was published in the peer-reviewed journal BMJ Quality and Safety on an open-access basis, so it’s free to read online.

Most of the UK media reports linked the research to the Department of Health’s plans to introduce new nursing associates, with some sources calling for the plans to be scrapped or reconsidered.

The headlines focused on the reported 21% increase in risk of death for patients if one qualified nurse was replaced by a less qualified assistant.

The media reports did not make it clear that the data about patient deaths applied only to patients who had undergone surgery.

Also, the limitations of the study, such as the potential for other confounding factors to influence results, such as doctor staffing levels, or local health policies, were not explained.

What kind of research was this?

This was a cross-sectional, observational study of nurses and patients from 243 hospitals, which also used mortality data for surgical patients from some of these hospitals.

Cross sectional studies can pick out associations between factors – in this case nursing skill mix and mortality, nurse survey data and patient survey data – but cannot prove that one causes another.

What did the research involve?

Researchers interviewed 13,077 nurses and 18,828 patients, and looked at discharge data for 275,519 surgical patients, from hospitals in Europe.

They asked patients and nurses about quality of care, and asked nurses about safety and how many professional and how many less qualified nursing staff were working on their last shift.

After adjusting their figures to account for factors that could affect the results, they analysed the data to see whether the mix of nursing staff in the hospitals was linked to mortality of patients who’d had surgery, and to patient and nurse ratings of quality and safety of care.

The 243 hospitals from Belgium, England, Finland, Ireland, Spain and Switzerland were part of a bigger Europe-wide study of nursing care.

Mortality was measured by the number of surgical patients in 188 of the hospitals (those with full data available) who died in hospital within 30 days of surgery.

Patients were said to have given hospitals low ratings if they described their care as anything less than excellent, or rated it as 8 or lower on a 10 point scale.

Researchers took account of the following confounding factors when making their calculations:

  • patient age and sex
  • emergency or routine admission
  • patient surgery type and other illnesses
  • total nurse staffing for the hospital
  • hospital size, teaching status and technology available

Researchers also checked if they needed to adjust figures for country-specific factors. 

What were the basic results?

The average staffing in hospitals in the survey was six care givers for every 25 patients, four of whom were professional nurses. However, this varied a lot between countries and hospitals.

There were on average 1.3 deaths for every 100 discharges from hospital after surgery.

The researchers found:

  • For every professional nurse replaced by a nurse assistant per 25 surgical patients, those patients have a 21% increased chance of death.
  • Every 10 point increase in the percentage of professional nurses (eg from 50% to 60%, or 60% to 70%) was linked to an 11% lower chance of death for surgical patients (odds ratio (OR) 0.89, 95% confidence interval (CI) 0.8 to 0.98).
  • Every 10 point increase in the percentage of professional nurses was linked to a 10% lower chance of the hospital being given a low patient rating (OR 0.90, 95% CI 0.81 to 0.99).
  • Every 10 point increase in the percentage of professional nurses was linked to a 15% lower chance of the hospital being given a poor safety rating by nurses (OR 0.85, 95% CI 0.73 to 0.99).

How did the researchers interpret the results?

The study authors said their research suggested “adding nursing associates and other categories of assistive nursing personnel without professional nurse qualifications may contribute to preventable deaths, erode quality and safety of hospital care.”

They say that any such policy initiatives should be taken with “caution” because “the consequences can be life threatening for patients.”

Conclusion

The headlines generated by this study are alarming, but there are some reasons to be cautious about the findings.

The study does not show that patients are more likely to die because of fewer professional nurses in the skill mix in a hospital. While that’s a possible explanation of the results, this type of study can’t tell us that for sure. It only tells us what happened at one particular point in time, not whether one factor led to another.

Other explanations – such as doctor staffing levels, or local health policies – might account for part or all of the findings. Researchers say they ruled out some explanations, such as whether hospital size or working environment had an effect, but this type of study cannot account for all possible explanations.

Also, some of the findings are close to the point at which they could be down to chance. The main finding of 10% lower chance of death, for example, has a margin of error that means the true figure could be anywhere between 2% and 20%.

One could also question some of the decisions around how the ratings were classified. Patients, for example, were considered to have given a low rating to their hospital if they said their care was anything other than excellent. It is likely that patients who said care was “very good” did not intend to give a low rating.

  • 1 Comment

Readers' comments (1)

  • As one of the study authors, I'd like to say that this is a generally fair piece that gives a balanced view of the study and its limitations. Well done NHS choices and to NT for republishing. Of course what it doesn't do is set this in the context of other available evidence which is really important when assessing studies like this.

    However, I would just note that where it says "....some of the findings are close to the point at which they could be down to chance. The main finding of 10% lower chance of death, for example, has a margin of error that means the true figure could be anywhere between 2% and 20%..." it is somewhat misleading.

    While statistical significance testing is a bit arbitrary it's the best we have. It gives a rule to answer the question "..is this likely to happen by chance when there is really no difference on average..". In this case the answer was no, it is NOT likely ddue to chance. The margin of error that is quoted answers a more important question - "...what do we think the difference is..." We are uncertain, the margin of error is wide. We think the difference is somewhere between 2 & 20% but it IS more likely to be somewhere near the middle... So while not absoutly wrong I think that this comment misses the point of the statisitcal analysis. It's fair to say that it is possible that in reality the negative 'effect' on mortlaity is much smaller than our estimate. It is equally fair to say it could also be much bigger!

    Unsuitable or offensive? Report this comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.