Unskilled personnel, skill mix, quality of care and patient safety in the healthcare services: a mapping review
Mapping review
|Published
This mapping review is a follow-up of a systematic review that did not identify any studies eligible for inclusion. We have therefore extended the inclusion criteria to include studies with weaker study design (e.g. observational studies).
Key message
In Norway, around 25% of personnel working in the municipal healthcare services lack a relevant health-related education (so called unskilled), while at the same time the care need of patients in the community is increasing. We know little about how this affects the quality of services and the safety of patients. In addition, we know little about the consequences of changes in skill mix, e.g. when registered nurses are replaced with lower qualified personnel. With skill mix we mean the composition of the various categories of nursing staff providing direct patient care. This mapping review is a follow-up of a systematic review that did not identify any studies eligible for inclusion. We have therefore extended the inclusion criteria to include studies with weaker study design (e.g. observational studies).
- We included 29 observational studies of the relationship between skill mix and patient safety outcomes, and four comparative studies, all of which evaluated different interventions.
- A majority of the studies were conducted in the United States. Only one study was conducted in Europe and none in a Nordic country.
- Skill mix measures, databases, risk adjustment variables, and terminology used to describe the assisting staff varied greatly between studies. In addition, the assisting staff’s degree of education or their work tasks were rarely described.
- The results were mixed for the vast majority of outcomes. However, a majority of observation studies reported a correlation between higher proportion of nurses and lower mortality rate. No study examined how a large proportion of unskilled personnel in the care team affects patient safety.
We still lack robust evidence of how skill mix (or a high proportion of unskilled personell) affects the quality and patient safety in the healthcare services. We found some support for an inverse relationship between the proportion of nurses and patient mortality, while the results for the other outcomes were mixed. It is uncertain if the results can be generalised to a Norwegian context.
Summary
Background
Nurse staffing, i.e. the number of nurses, nurse skill mix, and the experience and competence of nurses, are important factors to quality of care and patient safety. Optimally, staffing should match the care needs of different patient populations, to ensure that appropriate and safe care is delivered to all people, in all healthcare institutions.
There is some evidence from systematic reviews that lower nurse staffing (density) is related to more adverse patient outcomes. It is however less clear how various skill mix levels relate to the quality of care and patient safety, and very little is known about the effects of unskilled personnel working in direct patient care. There is a shortage of nurses both in Norway, and worldwide, and to recruit a sufficient number of qualified personnel is a challenge. This may increase the likelihood that more unskilled health workers are hired. In Norway, this problem is greatest in the municipal health and care services, where already a large number of unskilled personnel are at work. This is worrying partly because patients treated in municipal health services have an increasing need for care. The Directorate of Health has therefore commissioned this mapping review to get an overview of the research on this topic.
We have used McGillis Halls (2005) definition of skill mix which is "the combination of different categories of healthcare workers (read nursing personnel) that are employed for the provision of direct care to the patient." Higher skill mix is when proportionally more registered (higher qualified) nurses work in a healthcare team, and lower skill mix when the team consist of proportionally fewer registered nurses, and more licensed or unlicensed (assisting) personnel.
Objective
The aim of this follow-up report was to map:
1. Research on the relationship between skill mix, quality of care and patient safety in the health and care services. This includes studies of associations between the proportion of unskilled workers working in direct patient care and patient safety outcomes.
2. Research on the effect of professional mix on quality and patient safety in health and care services i.e. effects of higher vs. lower proportion of higher qualified care staff working directly with patients. This includes studies of the effect of unskilled workers working in direct patient care on patient safety outcomes.
Method
We conducted a mapping review as a follow-up of a previously published systematic review that did not identify any controlled studies that were eligible for inclusion. We extended the inclusion criteria to include observational studies (see new inclusion criteria below). We searched ten databases from April 2016 up to July 2017 for studies that evaluated either the relationship between, or the effect of, skill mix (including studies of unskilled personnel) on quality and patient safety in the healthcare services. The search had no language restrictions. In addition, we reviewed studies that were excluded after full-text screening (N = 179) from the previous search for the systematic review, and searched reference lists from studies that met the new inclusion criteria.
Two people independently screened all titles and abstracts from the update search, reassessed previously excluded full text studies, and reviewed possible eligible studies for inclusion in the review against the new revised inclusion criteria below:
|
Population: |
All healthcare services (both hospitals and nursing homes, etc.) |
|
Intervention: |
Higher/lower proportion of personnel without a relevant health-education, or of richer /less rich skill mix in a team |
|
Comparison: |
Higher vs. lower proportion of personnel without a relevant health education, or of highly skilled personnel (skill mix) in a healthcare team |
|
Outcomes: |
Outcomes related to quality of care and patient safety, e.g. information failure, malpractice, medication errors, abuse/violence and adverse outcomes (e.g. infections, pressure ulcers, falls, mortality) and patient satisfaction |
|
Study design:
Language: |
Randomised controlled trials (RCTs), non- randomised controlled trials (NRCTs), interrupted time series studies (ITS), controlled before-after studies (CBAs), cohort-studies, longitudinal studies and cross-sectional studies English, Norwegian, Swedish, Danish, German, Spanish and Icelandic. Languages that are not mastered by the project group will be considered if relevant language knowledge is available. Any references we cannot read will be listed in a separate table. |
We excluded studies that evaluated the effects of nurse intensity, i.e. studies that compared different nurse to patient ratios, or studies that only reported re-admissions, length of stay or costs, as this was not within the scope of this review. We solved disagreements through discussion between review authors.
Results
We found 29 observational studies that investigated the relationship between skill mix and adverse patient outcomes, and four comparative (non-randomised) studies evaluating the effect of changes in skill mix on quality and patient safety. We did not find any studies that evaluated the effects of having a high proportion of personnel without relevant healthcare-education working with patients in the healthcare services.
A majority of the studies were conducted in the United States and in other non-European English-speaking countries (Canada, Australia), and only one study was conducted in Europe. There was a large variation between studies in the type of skill mix measures, the databases and the risk adjustment variables used. The terminology used to describe the group of assisting personnel varied greatly between studies, and their degree of education and work tasks were rarely described.
The study results were generally mixed for most outcomes (i.e. for physical restraint use, medication errors, infections, pressure ulcers, patient falls). However, a majority of the observational studies provided support for a significant relationship between higher proportion of nurses and lower patient mortality.
Discussion
Most of the included studies were observational studies, which do not provide strong evidence about causal relations. The observational studies however gave some support for am inverse relationship between the proportion of nurses and patient mortality, but they showed for most other outcomes mixed results. Comparative studies can in principle identify causality. However, all four comparative studies evaluated different interventions, three of the four studies were relatively small, and in addition, they did not show consistent results.
The included studies used a number of different skill mix measures, a number of different data bases, and controlled for many different co-variables when analysing the data. Additionally, the studies used different terminology for the assisting personnel, and did not describe their degree of education or work tasks. All this makes comparisons between studies, and between countries difficult.
As a majority of the studies were conducted in the United States, and the American healthcare system is very different from the Norwegian, it is uncertain whether the results can be generalised to a Norwegian context.
Conclusion
We still lack robust evidence of how skill mix (or a high proportion of unskilled personell) affects the quality and patient safety in the healthcare services. We found some support for a relationship between higher proportion of nurses and lower patient mortality, while the results for the other outcomes were mixed. It is uncertain whether the results can be generalised to a Norwegian context, and there is therefore a great need for studies of skill mix performed in health systems relevant to Norwegian conditions.
