Technological innovation for workload allocation in nursing care management: an integrative review

Introduction

The nursing staff contribution is essential to guarantee quality and access to health services, as well as the outcomes of care for patients and their families. New technologies should support nursing management and decrease workload to meet the needs of the patients. This aspect is relevant since nursing care is essential for the patient recovery and safety (Moreno-Monsiváis et al. 2015).

According to the World Health Organization, there is a relationship between the nursing workload, the morbidity, and mortality of hospitalized patients (Ball et al., 2018) and after discharged (Kim et al., 2020). Nevertheless, the evidence is not strong to determine the best nurse-to-patient ratio and the effects on the nursing workload and patient outcomes (Coster et al., 2018).

The nursing workload is the amount of time, physical and cognitive effort needed to provide nursing care, in addition to activities related to service management and professional development (Pedroso et al., 2020). Excessive nursing workload goes against the values of humanization of care, patient outcomes, patient safety (Arango et al., 2015), quality of care (Romero-Massa et al., 2011; Chang et al., 2019), omitted care (Moreno-Monsiváis et al., 2015) and nurses’ health (Harvey et al., 2020). There is interest in creating technology to integrate scientific evidence and nursing expertise to reveal the relationship between nurse workload, burnout, and care quality (Farid et al., 2020). Thus, an interdisciplinary research team, including nurses and engineers, working together is required to understand this relationship.

The aim was to analyze the use of technologies on the distribution of workload for nursing care management. The following research questions (RQ) were addressed:

Methods

An integrative literature review was performed using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Studies were retrieved from SCOPUS, Scielo, PUBMED, and CINAHL in January 2021. The search strategy was: (“analytics” OR “operations research” OR digital* OR “IT” OR technolog*) AND (“nursing workload” OR “nurse-patient ratio” OR “nursing workforce planning” OR “nursing algorithm” OR “nursing personnel staffing and scheduling” OR “nursing staff”)). The authors were involved in the identification, screening, and analysis stages. Inclusion criteria considered studies that address the integration of technology on nursing workload, burnout and nursing acceptance, published in full text in English, Spanish or Portuguese classified as scientific articles or reviews, in the areas of nursing, medicine, social sciences, engineering, published between January 2016 and December 2020. This date range was considered to focus on the recent technologies. Articles related with technology that didn’t analyze the impact on nursing workload or didn’t present technologies were excluded.

For all research, this must include a final section including details of ethical approval, informed consent and, where relevant registration.

Initially 2,818 articles were potentially relevant, of these 21 that were duplicates were eliminated and 129 were discarded by automation tools because they didn’t meet the inclusion criteria. In the second phase, the titles of 2668 articles were screened and 2486 were excluded since they didn’t meet the inclusion criteria. Later, the abstracts of 182 articles were analyzed. In the third phase, two independent reviewers read the full texts and discarded 96 articles because they didn’t contribute to the RQ and 46 because they obtained a low score in the CCAT. Ultimately, 35 studies were included (Figure 1).

Figure 1. Flow chart of the study screening process.

Quality review of each article was independently reviewed by two members of the research team/authors using the CCAT, that has shown to have high reliability to rate research papers (Crowe et al., 2011). This tool is divided into eight categories and 22 items; each category receives a score on a 6-point scale from 0 (the lowest score) to 5. The total score is expressed as a percentage, and we arbitrarily considered acceptable a score above 60%. When there were differences in the quality assessment of an article, team consensus was achieved by comparing the arguments and evidence.

Data extraction was performed identifying the purpose, the design, participants and context of the study, the type of information collected, and the results. The interdisciplinary work between engineers and nurses made it possible to resolve doubts about technological and nursing concepts. For data synthesis, we performed an inductive content analysis. The information was collected in our database, and then grouped into categories that responded the RQ.

Results

Elements of nursing management

The studies were classified into three categories: 1. Nursing workload; 2. Information technologies and technological means for management; 3. Technology acceptance. Figure 2 presents a concept map of the categories discussed in the results.

Figure 2. Concept map of the studies’ categories and results.

• - Nursing workload

This category includes the methodologies for estimating nursing care requirements, factors, and approaches for calculating nursing workload and the predictors of nursing burnout and outcomes.

The challenge for workload estimation and prediction systems is having adequate information about the time and effort needed for nursing care activities, considering the interactions and iterations of the day-to-day care process, in physical, cognitive, organizational, and emotional dimensions (Redley et al., 2020). The variability in the patient care process usually has significant skewness. There are methodologies for estimating nursing care requirements, which are classified into those based on expert opinion, benchmarking, nurse-patient relationship, patient prototyping, multifactorial indicators, and task time-based approaches (Griffiths et al., 2020).

There is literature about nurse-to-patient ratios, but the validity and reliability of the studies is weak (Griffiths et al., 2020). In practice, legislation on minimum nurse staffing levels varies. In California (USA), law 394 of 1999 stipulates a 1:6 ratio. In Australia, the Safe Patient Care Act 2016 stipulates a 1:4 ratio. In Queensland, the daytime ratio is 1:4, and nighttime 1:7. In Wales (United Kingdom) the law stipulates a daytime ratio 1:5, and nighttime 1:7. Other countries, such as Colombia (South America), do not have legislation. The methodological proposals on how to calculate the nursing care requirements, nurse staffing, and skill mix, are based on local observations, which don’t allow generalization.

Low staffing levels negatively affect safety indicators and staff performance (Simonetti et al., 2020), but estimating the nursing workload, considering emergent care, for decision making, is still an unsolved challenge (Griffiths et al., 2020). There are no systems that estimate workload beyond ad hoc methodologies, based on average measurements that ignore the variability. The most currently used systems are:

• • GRASP (Grace-Reynolds Application and Study of PETO) (Clark & Poland, 1976).

• • TOSS (Time Oriented Score System dedicated to intensive care patients) (Italian Multicenter Group of ICU Research, 1991).

• • NAS (Nursing Activities Score dedicated to intensive care patients) (Miranda et al., 2003).

• • The Safer Nursing Care Tool SNCT (The Shelford Group, 2014).

• • RAFAELA Patient Classification System including the Oulu Patient Classification instrument and the Professional Assessment of Optimal Nursing Care Intensity Level (PAONCIL) (Aschan et al., 2009).

• • Safecare by Allocate Software.

Comparisons of these systems show significant differences in the results (Griffiths et al., 2020), and there is no evidence that one outperforms the others. In addition, the professional judgement-based approaches are popular and relatively reliable.

Commercial software computes staffing levels, coordinates nursing shifts, and plans nursing activities in case of unexpected staff absences. Tuominen et al. (2020) show that such tools reduce operating costs, the time required by nurse managers during the rescheduling process, the number of understaffed shifts, and unexpected changes in shift assignments.

The impact of technology for predicting nurse workload and deciding optimal staffing levels includes factors such as quality of care, patient safety, staff turnover and satisfaction, and process efficiency. Understanding nurses work is essential to assessing how technologies contribute to quality of care (Redley et al., 2020). There has been insufficient discussion of how technologies can be used to correlate operational cost with quality of care (Griffiths et al., 2020).

In relation to the predictors of nursing burnout and outcomes, patient and nursing staff outcomes are affected by nursing burnout and workload, which is significantly higher in ICUs than in other units (Youn et al., 2020). Burnout is a problem that originates in individuals and organizations (Maslach & Leiter, 2017). High nursing workload and poor skill mix are correlated with higher levels of burnout (Assaye et al., 2020; Simonetti et al., 2020), as does a poor work environment (Pekince & Aslan, 2020). In turn, burnout is associated with lower patient safety and omitted patient care (Dall’Ora et al., 2020).

High nurse workload is associated with higher rates of mortality (Assaye et al., 2020; Aiken et al., 2021), hospital-acquired infections, pressure ulcers, medication administration errors (de Oliveira et al., 2016; Kang et al., 2016), falls (Carlesi et al., 2017), treatment dropout (Assaye et al., 2020), increased length of stay (Magalhães et al., 2017) and readmissions (Assaye et al., 2020; Aiken et al., 2021). It is also related to needlestick injuries, intent to leave, and absenteeism (Assaye et al., 2020).

Harris et al. (2018) found a correlation between technology use and nursing burnout. The study showed that nurses experienced stress, burnout, and frustration with the use of EMRs and they reported insufficient amount of time for documentation.

The most studied technologies in ICUs (intensive care unit) were electronic medical records (EMR) and electronic medication management (Redley et al., 2020). Nursing-specific information systems, general computer, communication technologies and electronic data analysis being less often studied. Time savings are proven with the use of technology, but the impact on clinical decision making, interpersonal relationships, and work stress that underpin nursing care quality is unproven with respect to direct benefits. These systems could fail to record legitimate reasons for missing data, which can reduce patient safety (Hope et al., 2019).

• - Information technologies and technological means for management

Human and compassionate relationship between patients and nurses requires a meticulous synchronization between patients, healthcare staff, suppliers, processes, technologies, and information. The studies show that information technologies (IT) and technological means can support nursing care (Vorakulpipat et al., 2019; Zhou et al., 2019).

IT include EMR and systems for: online data input, tracking patients, detect changes in the patient condition and workload reallocation, trace patient’s calls, hands-free communication between nurses, detecting falls, barcode technologies for medicine management, data mining on nursing flows, among others (Massarweh et al., 2017; Respicio et al., 2018; Vorakulpipat et al., 2019; Yu et al., 2019; Zhou et al., 2019; Moon, 2019).

For effective IT implementation in nursing, change management and communication are determinants (Giraldo et al., 2018) as well as technology skills training (Farzandipour, 2020). It’s also recommended to follow workflows, multiplicity of task and patient care plans (Vorakulpipat et al., 2019).

Current technology to assess the nursing workload doesn’t capture the complexity or quality of care, and doesn’t recognize the concurrent, iterative, and interrelated nature of cognitive processes and activities in nursing. Therefore, there is a gap in the literature on how technology affects the quality of nursing care and how to measure it (Redley et al., 2020).

The patient’s condition has a significant effect on the workload (Sato et al., 2016). The development of early warning systems for patient deterioration (systolic blood pressure, pulse, respiratory rate, and oxygen saturation), using analytics, shows a significant decrease in nursing care time (Mau et al., 2019). These findings could be used to build a model for staffing (Leary et al., 2016).

• - Technology acceptance

The factors that influence technology acceptance are the perceive usefulness, the ease-of-use (Berg et al., 2017; Holden et al., 2016), and the participation of nurses in its design (Redley et al., 2019; Pendon et al., 2020). The human and organizational dimensions influence acceptance more than the technological dimensions (Handayani et al., 2018). The use of technology is beneficial when it promotes productivity and nurses’ empowerment, patient safety (Beaney et al., 2019), and quality of care (Higgins et al., 2017). Also, there is a positive correlation between the perception of usefulness and ease-of-use of EMR, the scanner to record the medication administration, and the use of lifts to mobilize patients and nurses’ technology acceptance (Tubaishat, 2018). These technologies contribute to reducing nursing staff fatigue (Steege & Dykstra, 2016) and nurses’ stress, while increasing their ability to focus on patient care (Ozan & Duman, 2020).

Tubaishat (2018) and Zhou et al. (2019) found that nurses age, sex, experience level and willingness to use the technology, are moderating variables of technology usage behavior, and predictors of perceived usability of EMR. On the other hand, the interface design of software functions plays a crucial role in technology implementation (Tu et al., 2018).

In contrast, the lack of confidence by professionals, past experiences, heavy workloads, and the time needed to learn are barriers to use technology (Wood et al., 2019). Additionally, nursing staff tend to use “informal workarounds” to deal with exceptional situations in administrative processes that have technology. These practices are a risk for patient safety (van der Veen et al., 2020) and information reliability.

Discussion

Technology in health services has benefits for streamlining processes, improving the quality of care, improving communication by supplying real-time information related to patient management and having data for decision making, process evaluation and results. The implementation success depends on the acceptance of technology.

Nurses have significant interaction with technology to monitor and manage patient care. Most studies look at the interaction of ICU nurses, but the technology requirement should extend to all health services, at all levels of complexity. Hence, an essential aspect in the implementation of technology is to enhance its acceptance (Phillips, 2019).

IT and care management systems in nursing should allow effective planning of nursing workload. These tools should promote patient safety, quality of care, and staff satisfaction. This is achieved when technology allows staff to spend more time with the patient and less time doing administrative processes. Although the usefulness of technology for nursing practice and clinical decision making has been demonstrated, the study on the impact of these is incipient and there is no evidence on the best way to plan workload allocation in nursing (Griffiths et al., 2020).

There are opportunities for research in this topic because the conclusions in the literature are contradictory. While some studies conclude that technology can increase staff productivity, empowerment, and improve nursing care outcomes, other’s show that technology can increase nursing staff stress and burnout. In addition, it is unknown how these technologies contribute to promote compassionate nursing care.

Further, the literature recognizes that studies analyzing the effect of technology use in nursing care is not free of bias (Redley et al., 2020). In these cases, although observational practices do not allow us to see cause-effect relationships, it is feasible to think of methodologies that resolve these questions (Griffiths et al., 2020).

Most of the software is limited because they don’t use real-time information. Instead, they work under the assumption that a sample is sufficient to generalize the needs of the patient population in the future. Furthermore, these solutions assume that once the contracted staffing level has been found, a fixed nurse-to-patient ratio and fixed number of hours will satisfy nursing care requirements (Griffiths et al., 2020). The strongest assumption relies on average care requirement times and allocating the load assuming that individual variations can be accommodated (Griffiths et al., 2020), ignoring the staff experience levels. Further, information systems should consider the scientific evidence showing that understaffing, low levels of skilled staffing, burnout, and high levels of intent to quit negatively affect clinical outcomes and increase turnover costs (Simonetti et al., 2020). Simultaneously, user-friendly systems are needed to allow nurse managers to make better decisions quickly and efficiently without creating staff burnout and stress (Harris et al., 2018).

The discipline of operational research techniques has proposed techniques to optimize staffing decisions using mathematical models or computer simulations, known as the “nurse rostering problem” and how to reassign staff to units with high demand. These algorithms have been able to prove that allocation based on averages may not be best in the face of variability in patient requirements and that this variability usually means more staffing to meet peak demand. However, how to calculate workload has not been discussed in depth (Griffiths et al., 2020), and how to use IT to estimate and predict workload including all the physical, cognitive, organizational, and emotional dimensions still is a major unsolved question.

Finally, in this research we have not taken into consideration sex and gender differences in the results since the reviewed papers do not clearly conclude on how the associated factors and barriers that promote the technology acceptance and usability is different for different sex and gender.

Conclusions

Based on the literature analysis, we conclude that IT has the potential to improve care management. There are several technological platforms that propose solutions to manage nurse workload and shift management. Further, the existing models use approaches based on expert opinion, benchmarking, nurse-patient relationship, patient prototyping, multifactorial indicators, and approaches based on task-based time requirements. These approaches differ importantly in how workload is assigned and the variables that are considered. It is equally important to analyze the technology and nursing workload in ICUs and non-critical units, but scientific evidence is more detailed in the former type of services.

The factors that promote the technology acceptance and usability of these models include the satisfaction for an increased time for direct care, a perception of improved patient safety, quality of care, and reduction of workload and staff burnout. The identified barriers for these technologies are the lack of knowledge of the technological tools, long times needed for their use, and low perceived ease of use. Thus, it is important to involve nursing professionals in the technological design process.

Our study reveals that most of the models and technologies are focused on intensive care units, and few studies for other types of care units. Also, all the considered studies are performed in developed economies, none of them are performed in economies in transition or developing economies. In addition, there is no scientific evidence to compare the technologies between each other and we found no scientific evidence showing improvements in care processes provided after using them. Further, with the proliferation of nursing workforce management systems, a review to evaluate efficacy is warranted. The main research gap that we identify is the lack of studies showing evidence of the impact of technology measured as the correlation between the nursing outcomes and staff well-being with the use of workload estimation and the strategies for planning workload allocation. One key contribution of this study is that the conclusions are obtained from the analysis of a multidisciplinary research team, including nurses with expertise in management and engineers with expertise in healthcare systems.

Summary