Essay: Healthcare quality and patient safety

Introduction
Healthcare quality and patient safety are top priorities within healthcare worldwide. Demographic and technological development have increased the requirements for health services and have put the healthcare system under pressure (1). Growing demands for advanced treatments, more complex nursing, high quality and safety, better patient satisfaction along with short hospital stays, high production, and reduced consumption of resources are some of the challenges the healthcare system is faced with (1). Healthcare policy priorities and management should be supported by research on organization and structure to contribute to a higher degree of evidence-based decision making in order to improve healthcare quality and efficiency (2;3).
An increasing body of research in health service evaluate how different organizational factors affect the healthcare quality, the outcomes or costs. Adequate nurse staffing has been found important for optimal health services and treatments, which leads to better quality, patient safety, and patient outcomes. Although this relation is convincing and has appeared in several studies during the past 30 years across different hospitals and countries (4-14), no overall consensus on a valid and reliable tool to assess the demand for nurses has been found. (15-17). Furthermore, most studies have been limited to retrospective studies or studies including data at hospital level. There is a lack of research at nursing unit level that explore the relationship of healthcare quality, nurse staffing and related organizational factors. The identification of organizational factors and the development, validation and tools to test reliability for the daily assessment of organizational factors can assist in predicting staffing needs to ensure a better distribution of the limited healthcare resources, when planning interventions to improve outcome. Despite the evidence on the relationship between adequate nurse staffing and patient outcomes, nurse staffing is in Denmark often determined on the basis of tradition, economy or simple calculations such as number of nurses per bed or per patient (18). No previous published studies on nurse staffing in Denmark were found.
Background
The Conceptual Framework
This study is guided by the of Quality healthcare model defined by Donabedian (19). The model included the three dimensions: structure, process, and outcomes and has been universally accepted and widely used. Donabedian postulated that each of the three dimensions is a necessary condition for the next. An optimal structural setting or organization in which healthcare takes place is a prerequisite for improved process of healthcare performed, which leads to better patient outcomes. Based on the model, studies of patient outcomes assess either association with process variables, or indirectly associations with structural factors that affects process variables that are measured or not measured (20). Nursing care are processes that will affect patient outcomes, but structural variables that support or hinder nursing care will also predict patient outcomes. Hence, it is not enough to focus on optimizing nursing care processes, optimal structural conditions are also necessary for optimal patient outcomes. Structural variables include a work environment with adequate nurse staffing levels. Donabedian´s Quality Model can contribute to a better understanding of prerequisites and conditions for improving patient outcomes. While studies have found a direct relationship between patient outcomes and nurses´ work environment and nurse staffing (2;3;21-24) it is understood that although not measured it is how staffing facilitate or limit nurses` ability to provide high quality nursing care.
Magnet hospitals
The Donabedian model has been used as the basis for solid research on Magnet hospitals in the United States. This research had emphasized the importance of structural factors for both process and outcome.
The research about Magnet hospitals can be traced back to the eighties (25;26). Due to a shortage of nurses, researchers examined why some hospitals were better at attracting and retaining skilled nurses (thus termed Magnet hospitals). A number of organizational conditions such as better communication, higher nurse staffing levels, a flat structure, a better work environment, and more educated nurses were characteristic for magnet hospitals as opposed to non-magnet hospitals (25) . Further, researchers also found better patient outcomes in Magnet hospitals compared to Non-Magnet hospitals (27). The better outcome in Magnet hospitals was ascribed to better nursing care provided (process), which again was explained by a better organizational support (structure) for nursing care. Magnet hospitals are regarded as prime examples of positive organizational conditions with excellent patient care quality, nurses with a high job satisfaction, and good recruitment and retention of nurses.
Missed nursing care
To further explore the link between structure and outcomes on the basis of Donabedian’s framework, Kalisch et al have developed a model of “missed nursing care” defined as “Needed nursing care, delayed, partially completed, or not completed” (28). In a study from 2011 Kalisch showed that low nurse staffing was a predictor for missed nursing care, when controlling for unit characteristics. These findings are the first ones to clarify, that the process of nursing care is hindered when nurse staffing is insufficient. Nurses have to coordinate, provide and evaluate interventions to manage patients’ symptoms and responses to treatment and care to promote health and healing for hospitalized patients. Kalish showed that when nurses have to prioritize required nursing care interventions, patients are exposed to a risk of delayed or lack of care, medication and treatment and complications as a result.
.
In a systematic review from 2015 including 42 studies, the prevalence of missed nursing appears to be high. Most nurses (55-98%) reported at least one missing required nursing care intervention from their last shift worked (29). However, the kind of missed care varied depending on the different measurements, most frequently missing interventions were related to teaching the patient, emotional care needs and surveillance activities. Patient outcomes that were significantly associated with missed care were: medication errors, patient falls, nosocomial infections, pressure ulcers, mortality, prolonged hospitalization and decreased patient satisfaction. Most frequently factors predicting missed care were staffing resource adequacy, the safety climate, nurse-reported care quality, and teamwork. Although no evidence from intervention studies have been found, the findings from these studies indicate that units with better nurse staffing have reduced missed nursing care and consequently better patient outcomes (29)
Nurse staffing and patient outcome
International research has exposed several organizational factors that influence patient outcome, nurse outcome and costs. Nurse staffing, nurses´ education level, experience, and work environment are important factors for patient safety, morbidity and mortality (7-15). In one of the most cited studies from 2002, Linda Aiken found that for each additional patient per nurse with a four patient caseload, patients´ likelihood of dying within 30 days after admission was increased by 7% (risk adjusted OR=1.07) (24). Several systematic reviews and meta-analysis of studies mostly from the USA and Canada support the relationship between nurse staffing and mortality (6;7;9;30),but in the last decade European countries have also contributed to the research area (31;32). In a recent European study RN4CAST from 2014 with nine participating countries and 300 hospitals including 422,730 patients, the effects of staffing and education were associated with mortality (adjusted for different factors) (14). The main results were similar to the previous US results that for one additional patient per nurse the 30 day mortality risk increased by 7% (OR= 1.07) and for every 10% decrease of nurses having a bachelor degree the mortality decreased by 7 %. (OR= 0.93).
Further, studies have reported preventable events associated with nurse staffing, however, the results are less consistent. Nurse sensitive outcomes associated to nurse staffing were falls, urinary tract infection pneumonia, wound infection and pressure ulcers (7;22;33-38). Patients were also reported to be more satisfied in units with better nurse staffing (39;40). In a Finnish study, the total patient satisfaction decreased when nurses in day shift had to care for more than eight patients (41).
Although, a large number of studies have described the relationship between nurse sensitive outcomes and nurse staffing, other studies only found a few significant associations or no associations between staffing and outcomes at all (42-44). Reasons for conflicting findings may lie in the methodological challenges within this research area. A review attempt to analyze explanations for inconsistencies from 29 reviews (45). Although the overall conclusion in the reviews was that a significant association between nurse staffing and patient outcomes exists, different definitions and methodological disparities were the main reason for the variations. A total of 82 different measurements of nurse staffing were identified across the primary studies, and 74 different patient outcomes that had been measured differently were identified. Variations in definitions and measurements might be explained by the data source, as most nurse staffing studies are based on data from administrative databases at hospital level with limited flexibility in deciding how variables are defined and measured (45). In addition, different analytical methods including risk adjustment strategies, together with different study designs (longitudinal vs. cross-sectional) were suggested as possible explanations for the inconsistency in results across studies. In another study, Mark et all found that increased nurse staffing was associated with patient outcomes, but that the degree of the effect depended on the general staffing level (10).
However, in relation to the large body of literature produced through the last 30 years showing a significant relationship between nurse staffing and patient outcome, the amount of studies that reveal inconsistency is limited. It can be concluded that a strong association exists between nurse staffing and different patient outcomes despite variations in definitions and measurements.
Nurse staffing and nursing outcome
Low levels of nurse staffing have also been associated with different adverse nurse outcomes, which again has been associated with patient outcomes. In a comparison of 30 hospitals in England, nurses were found to be 71% more likely to experience burnout and 92% were more likely to experience job dissatisfaction in hospitals with low nurse staffing levels (more than 12 patients per nurse) compared to hospitals with high staffing levels (up to 8 patients per nurse) (31). Similar results were found in a US study focusing on the relationship between inadequate staffing and nurse burnout and job dissatisfaction (46). For each additional patient per nurse the probability for reported burnout increased by 23% and for nurses’ job dissatisfaction the increase was 15%. In contrast, no association between perceived high staffing and more satisfied nurses were found in a US nurse survey from 2006 (47) .The opposing results were explained by different measurement (actual workload vs. perceived workload) and design (hospital level vs. nurse level).
Studies about staffing adequacy and nurses´ intension to leave their current job due to job dissatisfaction were also inconclusive. In the RN4CAST study, nurses with intension to leave their current job were estimated at 9% across 10 European countries, but no association to nurse staffing levels was found, although plans about leaving the job were associated with nurses burnout (48). In another study, higher rates of missed care were associated with more nurses intending to leave and with absenteeism, suggesting that better staffing might reduce intension to leave and absenteeism (49). Further, more medical errors and injuries as needle sticks have been linked to low staffing (50;51). In a review of the literature from 2004, low staffing levels were associated to burnout and needle stick, but nurses´ job satisfaction and absenteeism were not linked to staffing levels (6). Medical errors have also been linked to nurse staffing, showing that units with more nursing staff had significantly lower error rates (52-55). Although the overall pattern across studies found a beneficial effect of higher nurse staffing associated with less adverse nursing outcomes, the associations were not consistent due to possible methodical differences in studies or small effects that were difficult to prove.
Nurse staffing and costs
As nurses represent the largest clinical group in the hospital systems (56), a common reaction to cost containment is to reduce the number of nurses. However, appropriate staffing levels are essential for high quality care, including smooth patient admissions and well-planned discharges. Studies have revealed that reducing nurse staffing can have unintended negative clinical and financial consequences for patients, providers, and the healthcare system (57). A US study from 2013 concluded that patients with heart and lung diseases in hospitals with an average of eight nursing hours per patient had 25% lower odds for avoidable readmissions compared to hospitals with 5.1 nursing hours (58). In addition, studies have showed prolonged length of stay and more readmissions associated with low staffing, which may prove more costly than the saved nursing hours. Further, some studies have shown the relation between low staffing and increasing overtime, and higher nursing burn out, absenteeism and turnovers with more reductions in recruitment and orientation costs (59-65). Reduced nursing staff would cut costs, but to carry this out focus would have to be on attention to whether this would lead to unintended derivative financial consequences.
Mandated staffing
As a consequence of the numerous studies revealing an association between inadequate nurse staffing and higher rates of adverse patient and nurse outcomes leading to increased costs, a mandated minimum nurse-to-patient ratio has been introduced by legislation in California 2004. Different ratios for different unit types were introduced (66). For medical surgery units, a maximum of 5 patients per nurse at all times in acute care hospitals was set. Studies investigating the effect of mandated staffing agreed that legislation has increased the average nurse staffing level in California, but inconsistency was found in studies concerning the impact on patient outcome. A study comparing hospitals in California before and after the above legislation found that in hospitals that previously did not meet the standard, the mandated standard had the intended effect, but the better staffing did not relatively improve patient outcomes in general (67). This is similar to a review of 12 studies on mandating staffing in California. The review revealed no significant effect of improved staffing on nursing quality and patient safety. However, an increase in patients´ severity did not result in more adverse patient outcomes, which may reflect an effect of the mandated ratios to prevent more adverse outcomes (68).
A concern about hospitals hiring lower skilled nurses to meet the standard at lower cost was rebutted in a study finding California hospitals following the average national curve of increasing nursing skill mix (69). A qualitative study based on interviews with hospital directors in California have revealed the challenges in connection with honouring the mandated minimum for nurse staffing(70). Although hospital leaders reported that the legislation have had an effect on the overall patient service and nurse staffing, hospitals had to decrease the number of non-nursing staff in respond to financial challenges to absorb the increased cost of more nurses. The missing effect on quality might partly be explained by nurses forced to do non-nursing tasks. Further, instead of the intention to secure a maximum level of patients per nurse, the inflexible standard today is to add more staffing when acuity changes. The hospital leaders further report that patient satisfaction surveys have not changed following the implementation of staffing ratios, and nurses’ experiences seem to be mixed. Less patients to care for has reduced burnout and brought staffing up to a needed level, especially in hospitals with low staffing prior to the legislation, but nurses experience less autonomy and more ancillary service responsibilities (70).
A study that found a positive effect of the legislation had compared hospitals in California to hospitals in Pennsylvania and New Jersey (66). Nurses in California cared for on average two fewer patients on medical and surgical units and reported a consistently better quality of care and had lower rate of burnout and job dissatisfaction. In addition, better staffing was associated with significantly lower mortality. The results from California are followed closely in other states considering similar approaches to improve patient care.
Due to difficulties in maintaining nurses in hospitals, mandated minimum nurse-to-patient ratios have been legislated in Victoria, Australia since 2001 (71). A more flexible model than the one in California was enacted in Australia with a minimum nursing staff at unit level of 5 nurses for 20 patients in medical/surgical unit level 1. Compared to California, this enables the possibility of 5 patients per nurse during a shift, while other nurses take care of only 3 patients depending on required patient care. Although mandated minimum staffing has been effective to attract nurses back to hospitals, no published empirical studies from Victoria were found to support the effect on healthcare quality and patient outcome.
Concept of nurse staffing
In the identified studies, nurse staffing is most frequently associated with patient, nurse, or economical outcomes. No published studies have demonstrated an inverse association (better staffing – poorer outcomes) and a few studies showed no significant association ( ). Although some results are varying, the majority of the studies describe a significant association between high nurse staffing and better outcomes. As mentioned previously, a lot of different definitions and measurements have been used to estimate nurse staffing, and several suggestions for future research in staffing have been made. These recommendations have to be taking into account and considered in the effort to get closer to a concept of nurse staffing.
Previous studies on nurse staffing have been criticized for not including factors affecting staffing and the provision of care when associations were explored. Arguments such as nurse staffing that goes beyond numbers sound reasonable (72) + ref 74 save staffing. Merely increasing the numbers of nurses might have limited effect if the nurses do not have sufficient education and experience (73).
A definition from 1975 of the term nurse staffing is: “The numbers and kinds of personnel required to provide patient care to patient or client”(74). This definition includes the number, but also the kind of nursing personal, but additional factors might be considered when exploring nurse staffing. In the publication “Safe Staffing save lives” from the International Collaboration of Nurses stated: “Safe staffing means that an appropriate number of staff, with a suitable mix of skill levels, is available at all times to ensure that patient care needs are met and that hazard free working conditions are maintained.” (74). In this definition factors affecting staffing are: number of staff, skills of the nursing staff, patient care needs and working conditions. Similar dimension of staffing has been recommended in a study from 2007 called “Looking beyond numbers”. The study stated: “An effective staffing model must take into account such variables as patient acuity, unit layout and ancillary support in determining the appropriate number, skills, experience, specialized training and education of nurses on a given unit” (72). Further, the “White Paper” on nurse staffing from the American Nurse Association was cited: “Models that consider additional variables that more closely match patient need with professional skill mix, experience, and the conditions under which nurses provide care, offer the precision necessary for today’s complex healthcare environment and patient needs” (75). To facilitate an understanding of how appropriate nurse staffing levels contribute to patient outcome a concept called” Nurse Dose” was developed in 2006 (73). The model consists of three essential components derived from the literature to assess nurse staffing: Dose (number of nurses or the amount of nursing care), nurses’ characteristics (education, expertise, and experience), and host (organization). The concept “Nurse Dose” has in two later studies been revised and tested in acute care settings to evaluate the association with patient outcomes (76;77).
The concept of nurse staffing contains a number of suggested organizational factors that alone or together contribute to ensure “save staffing that saves lives”. Based on that a nurse staffing model for this study was developed.
Nurse staffing model
There is no international consensus on a model to evaluate or determine adequate nurse staffing. Several factors affecting nurse staffing have been suggested to be included in staffing models to determine adequate nurse staffing and achieve high healthcare quality and ultimately secure better patient, nurse and financial outcomes. The potential relationship between the introduced factors affecting healthcare quality and staffing adequacy are summarized in Figure 1, including the numbers of nurses and patients, nurse competences, nursing intensity (patients` care needs) and the work environment.
Figure 1. Nurse staffing model including factors affecting healthcare quality and nurse staffing adequacy.
Number of nurses and patients refer to a relation often expressed by the number of patients per nurse or by available nursing hours per patient. Nurse competences include educational, experiential and expertise-related nursing resources, while nursing intensity is defined by the amount of patient care needs. Number of nurses and nurse competences in relation to nursing intensity can be regarded as a fraction between resources and needs. The nurses´ work environment contains all aspects affecting this fraction or balance, positively as well as negatively. From Magnet hospital research, a supporting work environment include for example: better communication, collaboration, supportive service systems, good managers and leaders, and an appropriate organization (25).
Further, the three organizational factors: number of nursing and patients, nurses´ competences and nursing intensity are in the model placed in a triangle to illustrate they are interconnected. For instance, if an increase in nursing intensity occurs (e.g. more sick/ depended patients), additional nursing staff and/or higher competences will be needed to reach an appropriate balance between resources and demands. Aspects in the work environment affecting the balance are illustrated by a surrounding circle in the model.
An inappropriate balance may affect nurses’ perception of staffing adequacy negatively, but the question is, at which point the organizational factors will affect nurses` ratings of staffing to be high or low. Further, a high level of staffing resources might not necessarily lead to improved healthcare quality. Consequently, nurses´ perceptions of healthcare quality were added in the model to explore preconditions for nurses with a view to secure high healthcare quality leading to optimal patient outcomes. Several definitions of healthcare quality exist from different perspectives (78). From a providers´ perspective the definition from the Agency for Healthcare Research and Quality (AHRQ) is useful and quite easy to deal with. Healthcare quality is defined as “Doing the right thing, at the right time, in the right way, for the right person – and having the best possible results”. Hence, high healthcare quality means that patients are getting the healthcare service they need, when they need it, using the best care interventions from the right skilled providers to get the best possible results (79).. Nurses might be well suited to perceive the healthcare quality, as they are around the patient 24 hours per day. Further description of the organizational factors, staffing adequacy and healthcare quality can be found in the method section.
Nurse staffing can be determined from a macro level (number of nurses per inhabitants) to a micro level (available nursing hours for a single patient). Most studies compare nurse staffing at macro level between hospitals with aggregated data from either administrative databases and/or from nurses` surveys. This allows for use of large amounts of data to identify small significant differences that contribute to important overall knowledge about the influence of different staffing levels on quality of care and patient, nurse, and financial outcomes. There is a lack of published studies evaluating nurse staffing at the unit level. Guided by the nurse staffing model this study contributes to our understanding of the relationship between organizational factors and healthcare quality and nurse staffing adequacy at the unit level using daily collected data.
Aims
The overall aim of this PhD study was to identify, develop, validate and reliability-test tools for the assessment of organizational factors related to nurse staffing at unit level, and subsequently to examine the association between these factors and nurses’ perception of staffing adequacy and healthcare quality. The PhD study has been based on a literature review and the aims of three related articles:
• To identify tools for the assessment of nurse staffing, nurse competences, the work environment, nursing intensity, and nurses’ perception of staffing and healthcare quality.
• To develop, implement, validate and reliability test a tool that measures nursing intensity by estimating time spent on nursing care interventions.
• To translate, validate and reliability test the Practice Environment Scale of the Nursing Work Index (PES-NWI) in a Danish context and to compare Danish nurses’ ratings of their work environments to Magnet and non-Magnet hospitals in the USA.
• To examine the association between nurse staffing (nursing hours per patient, nursing skill mix (percentage registered nurses), experience, expertise, and nursing intensity) and nurses’ perceptions of staffing adequacy and healthcare quality measured at the shift level