Nursing Assistant Staffing Model: Implementation and Outcome Evaluation in Taiwan Hospitals

ABSTRACT Background A key policy of the Taiwan healthcare system is promoting the hiring of nursing assistants (NAs) to reduce the workload of nurses. However, few studies in the literature have compared the relative effectiveness of different nurse staffing policies, and no studies have addressed the impact of the NA staffing (NAS) model. Purpose This study explored the current status of the NAS model as implemented in Taiwan hospitals and to investigate the methods used to evaluate the outcomes of NAS model implementation. Methods A cross-sectional online survey was performed in hospitals nationwide from June 1 to December 31, 2018. There were 153 government-accredited hospitals being invited to participate in the survey. Results Of the 139 hospitals that completed the online survey, 26% (36 hospitals) had implemented the NAS model. Many of the hospitals had hired full-time, certified NAs. Most of the tasks assigned to NAs were in the nonprofessional/general and technical work categories. In addition, medical wards were the most frequently assigned workplace, especially on day shifts. Few of the surveyed hospitals had evaluated the effectiveness of the NAS model. Improvement after implementation of the NAS model (based on the 35 outcome indicators) was found to vary significantly. The most commonly used outcome indicator was patient satisfaction. All the stakeholders, including patients, family caregivers, nurses, and nurse supervisors, supported implementation of the NAS model. However, some of the participants expressed concerns regarding the difficulties involved with model implementation. Conclusions/Implications for Practice Although the NAS model may be beneficial, further empirical research is necessary to confirm the contributions of NAs. Additional evidence regarding the benefits of the NAS model for hospital performance and patient outcomes may be expected to motivate more hospital administrators to adopt this model.


Increased Workload and Shortage of Nurses
A chronic shortage of nursing staff distresses many healthcare institutions around the world. In Taiwan, implementing a managed care-based national health insurance system in the 1990s significantly changed medical care patterns in the country, leading to shorter hospital stays, rising numbers of patients with multiple comorbidities who require relatively complex hospital care, and rising numbers of patients who are reluctant to be discharged because of insufficient self-confidence in their self-care ability. Constant workload upsurges and work-related stress have contributed to high turnover rates in the nursing workforce in recent years (Yeh et al., 2019). As of September 2020, of the 178,709 registered nurses (RNs) in Taiwan, only 64% were currently practicing. In contrast, the RN practice rate in Canada and the United States during the same year was 93.6% and 83.2%, respectively (Taiwan Union of Nurses Association, 2020). Despite the efforts of hospitals to recruit and train new nurses, staff shortages continue to destabilize the workforce and negatively affect healthcare quality (F. Pang et al., 2019). A multidimensional solution to the nursing workforce shortage is required. The World Health Organization (WHO) introduced the positive practice environment campaign in 2010 to raise public awareness about the need to improve the working conditions and environments of healthcare workers. Recently, the WHO has advocated for the improvement of healthcare policies and workplace safety for nurses and for increased support for nursing education and employment (WHO, 2020). In Taiwan, strategies to help ease the nursing shortage have been proposed by public, academic, and healthcare organizations (Huang & Lu, 2017;Ministry of Health and Welfare, Taiwan, 2018;Yang & Chen, 2018).

The Delegation Solution: Mixed RN/NA Staffing Model
One solution proposed to ease the problem of overworked nurses is to delegate the nonprofessional responsibilities of nurses to nursing assistants (NAs) and other nonnursing professionals. In the United States, many nursing tasks that do not require professional expertise are already regularly delegated to unlicensed hospital personnel. A nationwide survey in the United States found unlicensed hospital personnel (ranked from most to least working hours) to include NAs, patient care technicians, surgical technicians, transporters, medical assistants, surgical aides, monitor technicians, and unlicensed graduate nurses (S. Li et al., 2017). Similarly, NAs in Taiwan are mainly responsible for general tasks that do not require a high level of skill or expertise, such as processing hospital admissions and providing physical help to patients (Huang & Lu, 2017;Pang et al., 2019;Yang & Chen, 2018). Employing unlicensed personnel lets professional nurses allocate more of their time to direct care, which pushes forward patient safety and healthcare quality (S. Li et al., 2017).
Studies have found that a mixed staffing model in which RNs and NAs work as a team can deliver better care quality than the standard nurse staffing model if the ratio of RNs to NAs is sufficiently high (Aiken et al., 2017;Simonetti et al., 2020). In Europe and the United States, many medical institutions have excessively low RN-NA ratios because of reduced RN numbers to minimize costs. Low RN-NA ratios have been linked to higher rates of adverse outcomes, such as preventable deaths and reduced quality of care and patient safety (Aiken et al., 2017;Griffiths et al., 2019).
However, the application of NAs in Taiwan differs significantly from other countries. A national survey discloses that 299 hospitals in Taiwan employ only 8,017 NAs (Y.-C. Chen & Lin, 2014). Taiwan's RN-NA mixed staffing model thus clearly differs from the skill-mixed care model used in Western countries. Although not a substitute for RNs, NAs are assigned to the ward according to established delegation guidelines (e.g., American Nurses Association & National Council of State Boards of Nursing, 2019) and perform general duties for nurses. A mix of RNs and NAs working together in a single shift is referred to in this study as the NA staffing (NAS) model. The tasks performed by NAs vary according to the job description of each hospital and may include body cleaning, excreta disposal, feeding, medication, daily activity, patient safety assurance, ward maintenance, garbage removal and sorting, and other similar tasks (Wu & Wang, 2017;Yang & Chen, 2018). In this study, NA tasks are classified into the two categories of technical tasks related to patient care and nonprofessional general tasks related to daily logistics (Y.-M. Chen et al., 2016).

Impact of the NAS Model and the Research Purpose
The mixed RN-NA staffing model showed to reduce both the nonprofessional workload and overtime hours of RNs (Deravin et al., 2017;Jenkins & Joyner, 2013;H.-C. Li et al., 2014). Therefore, NA-supported RNs report higher job satisfaction (Feng et al., 2008). In a single-hospital study performed in Taiwan (Yang & Chen, 2018), most of the surveyed RNs agreed that integrating NAs into the staffing model was appropriate and worth implementing in acute care settings throughout the hospital. However, 40% of those surveyed stated that the NAS model did not reduce their overtime hours. Thus, a significant minority of RNs did not perceive that the NAS model had achieved one of its primary objectives. Although some hospitals in Taiwan have hired and trained personal care attendants to reduce the RN workload and maintain healthcare quality (Yang et al., 2015), the role of NAs has yet to be clearly defined and no universal indicators for evaluating NA performance have yet been established (Cheng et al., 2017;Liou & Chang, 2007;Yang & Chen, 2018). Nursing administrators throughout Taiwan have implemented the NAS model as an innovative approach to reducing RN workload. Therefore, the aim of this research was to investigate Taiwan hospitals that have implemented the NAS model and determine how they evaluate the effectiveness and impact of this model.

Methods
A cross-sectional survey was performed from June 1 to December 31, 2018. Calculations performed using G*Power 3.1.9.4 revealed that achieving a power of 80%, an alpha of .05, and an effect size of .25 required a sample size of 102 subjects. We selected hospitals accredited by the Taiwan government for research at all three hospital levels and included 21 medical centers, 82 regional hospitals, and 50 district hospitals. In the selected hospitals, nurse managers responsible for supervising NAs were invited to complete an online survey. Nurse managers in 139 of the 153 selected hospitals completed the online questionnaire (response rate: 90.8%). The characteristics of the surveyed hospitals, including hospital level, business type, geographic area, and NAS model implementation status, are shown in Table 1.
The survey questionnaire comprised 18 multiple-choice items and three open-ended question items (T.-Y. Wu & Wang, 2017;Yang & Chen, 2018). Three nursing supervisors and two senior nurses reviewed the questionnaire to ensure that the survey items were relevant and clear. The content validity index was .91. Two head nurses with experience managing NAs assisted with the pilot test of the questionnaire. The information collected using the survey questionnaire included background information and personal characteristics of the participants, whether the NAS model had been implemented, and NA personnel qualifications (i.e., workforce sources, educational level, and training certificate) and working conditions (i.e., workforce allocation and workload). The questionnaire also surveyed the frequency and type of auxiliary tasks assigned to NAs. To provide a reference for participants, the survey form listed 34 auxiliary tasks (24 technical tasks and 10 nonprofessional general tasks) typically performed by NAs, according to previous research (Y.-M. Chen et al., 2016;Wu & Wang, 2017;Yang & Chen, 2018). The participants used a 5-point Likert scale to specify how frequently they assigned NAs to perform auxiliary tasks (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always).
Other questionnaire items were used to collect information about the indicators used at their institutions to evaluate NAS model effectiveness, their perception regarding whether the indicators improved model effectiveness, and how their institutions evaluated NA performance. After referencing other studies that evaluated work outcomes of NAs, we divided the reported indicators into three categories: healthcare quality related, NA related, and RN related (Campbell et al., 2020;Fernandez et al., 2012;Pang et al., 2019). On the questionnaire, the participants rated their satisfaction with NA work performance using a 5-point Likert scale ranging from very satisfied (5) to very dissatisfied (1). The participants were asked to rate NA performance in terms of their caregiving behaviors, interpersonal relationships and communication skills, and overall satisfaction from multiple perspectives, including patient/family, nurse, and nurse supervisor. Finally, the questionnaire included open-ended questions that queried their opinions regarding the advantages/ disadvantages of the NAS model and suggestions for improving model implementation.

Statistical Analysis and Ethical Considerations
The JMP 13.0 software package (JMP Statistical Discovery LLC, Cary, North Carolina, USA) was used for the statistical analysis work in this study. Data were analyzed using descriptive statistics, one-way analysis of variance, and the Tukey-Kramer post hoc test. Ethical approval for this study was attained from the institutional review board at the authors' institution (IRB-EII-20180031). After completing the consent form, the participants completed the online survey anonymously.

Current Implementation and Work Conditions of NAs
Only 36 of 139 (25.9%) hospitals surveyed had applied the NAS model, including two medical centers, 16 regional hospitals, and 18 district hospitals. As shown in Table 1, the NAS model was most frequently applied in district and regional hospitals, most of which were private hospitals in southern Taiwan. The motives given for implementing the NAS model included reducing the workload and work stress of RNs (100%), improving RN job satisfaction (77.8%), reducing RN overtime work (75.0%), increasing patient and family caregiver satisfaction (72.2%), and ensuring quality of care and patient safety (72.2%).
At the 36 NAS model hospitals, most NAs were assigned to hospital wards, including internal medicine (91.7%), surgery (55.6%), obstetrics and gynecology (19.4%), and pediatrics (19.4%). The working conditions of the NAs varied. NAs worked 8-hour day shifts at two medical centers (100%), 15 regional hospitals (93.8%), and 10 district hospitals (55.6%) and worked 12-hour shifts at several district hospitals. Although NAs did not work night shifts at medical centers, they worked evening shifts (58.3%) and night shifts (36.1%) at some regional and district hospitals. The NAs assigned to wards worked as functional assistants for all RNs in the same shift. Most (61.1%) of the hospitals assigned one NA per shift. Relatively few of the regional hospitals and district hospitals assigned two (8.3%) or three (5.6%) NAs per shift. Thus, the number of patients to whom NAs provided care varied widely (range: 30-60).

Workforce Quality and Task Allocation
As shown in Table 2, most (77.8%) of the surveyed hospitals had recruited their NAs and most (52.7%) offered full-time positions. In a few hospitals (16.6%), NA personnel were contract workers outsourced from human resource agencies. In most of the surveyed hospitals, retention was higher among full-time NA personnel who had been recruited by the hospitals than among those contracted through human resource agencies. Some of the surveyed NAs had worked at their hospital for 10 years. To ensure NA staff work quality, most (72.2%) of the hospitals required training certificates as part of the hiring had not yet passed their licensing examination required applicants to present their nursing school diplomas during interviews (11.1%).
The study classified the work assigned to NAs into two categories: nonprofessional general tasks and technical tasks. One-way analysis of variance and post hoc tests were performed to compare the frequencies of work assignments among the three hospital types. In medical centers and regional hospitals, the frequency of nonprofessional general tasks performed by NAs (4.2 AE 0.5 and 4.0 AE 0.5, respectively) was significantly higher than that of technical tasks (2.9 AE 0.0 and 3.2 AE 1.0, respectively). In district hospitals, the frequency of nonprofessional general tasks performed by NAs (3.5 AE 1.0) did not significantly differ from that of technical tasks (3.5 AE 1.1).

Evaluation of the NAS Model
Outcome evaluations had been performed in less than half (41.7%) of the NAS model hospitals (two medical centers, seven regional hospitals, and six district hospitals). The number of outcome indicators used to evaluate the NAS model was one indicator in seven hospitals (19.4%), two indicators in two hospitals (5.6%), three indicators in two hospitals (5.6%), and four or more indicators in four hospitals (11.1%). Then, the 35 reported indicators were divided into three outcome categories, with 16 indicators associated with healthcare quality, 10 associated with NAs, and nine associated with RNs (see Table 3).
To compare the changes resulting from implementation of the NAS model, the participants rated the positive change in each indicator on a scale ranging from 0 (no improvement) to 7 (complete improvement). As shown in Table 3, 29 (83%) of the 35 indicators were perceived as improved (i.e., average score > 4). Perceived improvements in healthcare-quality-related outcomes were smaller compared with those in outcomes related to RNs and NAs. For healthcare-quality-related outcomes, all hospitals reported improvements in injury-prevention indicators, for example, accurate pressure injury care and the use of protective restraints. However, the smallest reported improvements were found in patient satisfaction, customer complaints, and patient safety incidents (e.g., falls and displacement of catheters

The Journal of Nursing Research
Pei-Hsuan YANG et al.
or drainage tubes). One hospital reported a decreased rate of companionship in its hospitalized patients. In Taiwan, hospitalized patients are usually accompanied by caregivers, who are typically family members or care agency personnel. In hospitals that involve NAs in the caregiving process, family members are relieved of their patient accompaniment duties. Regarding outcome indicators related to NAs, improvements were found in all the indicators, including job satisfaction, work performance, staffing hours, and certificate upgrade. Improvements were found in multiple indicators of RN-related outcomes, including job satisfaction, retention rate, overtime hours, and participation in training programs. However, overtime hours and participation in training programs were found to have only partially improved. As shown in Table 4, the average score for overall satisfaction with the NAS model was 4.0, signifying that the nurse supervisors were satisfied with both the NAS model and NA performance. Satisfaction scores did not significantly differ by hospital level. The level of satisfaction among patients and family caregivers with NA caregiving behaviors was the highest in medical centers, followed by district hospitals and
regional hospitals (F = 4.57, p = .002). According to the Tukey-Kramer post hoc test results, level of satisfaction among patients and family caregivers with the caregiving behaviors of NAs was significantly higher in medical centers (mean = 5, SD = 0.0) than regional hospitals (mean = 3.9, SD = 0.5). The qualitative data attained using the open-ended questions in the survey highlighted the advantages and disadvantages of the NAS model. A positive outcome noted for the NAS model was the increase in time available for nurses to provide professional nursing care and psychological support to patients. However, some of the participants conveyed concerns about higher personnel costs at the organizational level and about excessive reliance on NAs at the individual level.

Discussion
Hospitals in many countries have implemented the mixed staffing model in recent decades to compensate for acute shortages of RNs. The WHO has developed guidelines for effectively implementing a mixed staffing model in the healthcare workforce (Buchan et al., 2000). The advantages of integrating RNs and NAs in a mixed staffing model are well recognized in hospitals in Taiwan (Feng et al., 2008;Sun & Yen, 2010;Yang & Chen, 2018). However, an unexpected outcome of this investigation was that the NAS model has not yet been widely implemented in Taiwan. Only 25.9% of the surveyed hospitals employed NAs, with motives for implementation including reducing RN workload and improving RN job satisfaction. Although implementing the NAS model grasps positive results, role conflicts between RNs and NAs and poor teamwork are commonly cited problems (Buchan et al., 2000;Conway & Kearin, 2007;S.-C. Lin et al., 2019;Pang et al., 2019;Yang et al., 2015). The open-ended questions in our survey revealed financial issues (i.e., increased salary costs, increased costs related to managing NAs) to be the primary concern of the participants in implementing a mixed staffing model. Thus, increased costs may be the main reason the NAS model is rarely implemented in Taiwan hospitals. Historically, the roles and functions of healthcare professionals have evolved to meet changing public needs. A recent study exploring changes in staffing skill levels at hospitals in the United States during 2010-2015 (Pittman et al., 2018) identified a decrease over time in the staffing ratio of high-and intermediate-skilled NAs and an increase over time in the staffing ratio of lower-skilled NAs and RNs. The skill level of healthcare personnel should be appropriate to meet patient needs. Therefore, when determining the applicability of the NAS model, staffing type and quality are essential considerations. Several respondents in this study expressed concerns that low qualifications for NAs could compromise healthcare quality at their institutions. Although an effective NAworkforce requires appropriate education and training, Taiwan has not yet issued guidance on NA qualifications. Studies emphasize that, to be effective, NAs should have at least a high school education (B.-Y. Chang & Wang, 2013;L.-W. Lin et al., 2003) The results of this research revealed that, because of difficulties encountered in recruiting NAs, some Taiwan hospitals are willing to hire NAs without a high school diploma. Like other studies, this study found caregiver certification to be an essential qualification for NAs (L.-W. Lin et al., 2003;Pang et al., 2019;Yang et al., 2015), as differences in education and training significantly influence NA job performance. Lack of standardized education and training for NAs brings about inconsistent care quality and may increase RN workload, as RNs may assume the added burdens of training and supervising NAs (B.-Y. Chang &Wang, 2013;Lee et al., 2005). To improve NA quality, the Taiwan Ministry of Health and Welfare recommends hospitals employ nursing graduates who are not yet licensed nurses (Ministry of Health and Welfare, Taiwan, 2018). In support of this, C.-W. Chang (2013) found that, in terms of providing auxiliary personnel, unlicensed nursing graduates outperformed certified NAs in performing common NA tasks in hospitals. With the statistics of over 10,000 unlicensed nursing graduates each year in Taiwan, the NA workforce should be attainable. Therefore, to ensure high-quality healthcare personnel and to meet the healthcare needs of patients, the education and healthcare authorities should cooperate in restructuring the healthcare workforce to create the best human resources.
The healthcare systems of the United States, Australia, and some Asian countries have established authorization and supervision mechanisms for NAs and have clearly defined NA roles and functions (American Nurses Association, 1997; Y.-M. Chen et al., 2016). This study revealed that NAs in medical centers and regional hospitals perform significantly more nonprofessional general tasks than technical tasks. In acute care wards in Taiwan hospitals, especially those in medical centers and regional hospitals, patients are typically in critical condition, patient turnover rates are high, and the workload of auxiliary inpatient services is heavy. General nonprofessional tasks performed by NAs include changing sheets, moving patients, introducing the hospital environment at admission, and arranging the ward at discharge. However, the specific work performed by NAs may vary according to patient need. Many studies report that NAs often perform assistive technical tasks, such as bathing patients, measuring vital signs, and providing food and medicine through gastric tubes (Cheng et al., 2017;Pang et al., 2019;Yang & Chen, 2018). To reduce RN workload and improve nursing care quality, the work responsibilities of NAs should vary to meet the specific nursing needs and safety goals of clinical patients. However, hospitals must ensure that effective training is arranged for NAs to maintain good care quality.
The ideal way to evaluate the effectiveness of the mixed staffing model is to measure its impact on patient safety and quality measures. However, the difficulty in doing this is identifying indicators affected only by the staffing model. In Taiwan hospitals, indicators used to evaluate the effectiveness of the mixed staffing model include length of stay, infections, falls, and labor costs (Feng et al., 2008;Pang et al., 2019;Yang et al., 2015). This study found that fewer than half of the hospitals in Taiwan have implemented the mixed staffing model and that those hospitals use a variety of indicators to assess model effectiveness.

The Journal of Nursing Research
Pei-Hsuan YANG et al.
The most widely used indicator was patient satisfaction, which is also commonly used as an overall healthcare quality indicator (Currie et al., 2005). Previous studies have reported that both patients and healthcare professionals believe NAs contribute positively to healthcare quality (Yang & Chen, 2018). This research, which evaluated satisfaction from the perspective of stakeholders, including patients, family caregivers, nurses, and nursing supervisors, revealed similarly positive assessments of the NA role. Because the effectiveness of the NAS model requires an appropriate allocation of NA staff and RN staff and an appropriate assignment of work responsibilities to NAs and RNs, a sound human resource policy is crucial to achieving a situation that concurrently meets the needs of hospitals, nursing staff, and patients.
Finally, the main motivation for hospitals to employ NAs was identified as reducing the workload of RNs. The RNs in the surveyed hospitals worked all shifts, whereas NAs mostly worked day shifts and only occasionally worked evening and night shifts. Obviously, there is a gap between demand and actual workforce allocation.
A limitation of this study was its small sample size, which limited our ability to investigate the extent to which the NAS model reduced RN workload and affected healthcare quality. Future studies should collect more complete data for an objective assessment of the effectiveness of the NAS model based on multifaceted indicators. To persuade hospital administrators to apply the NAS model, additional evidence-based research on patient outcomes and organizational outcomes will be needed.

Conclusions and Implications
Recent societal changes in Taiwan present many near-term challenges for hospitals. These include increasing healthcare demand by population aging and tighter labor supply because of chronically low birth rates. Owing to the essential role of nurses as frontline healthcare workers, an effective nurse staffing model is needed to help manage their workload and maintain their efficacy. Although the Taiwan government has encouraged domestic hospitals to adopt the NAS model since 2014, few hospitals have done so. Obstacles to adopting the NAS model include inflexible human resource policies, lack of unified NA qualifications, and poorly distributed work responsibilities among assistive personnel. In addition, outcome indicators must be clearly defined to assess the effectiveness of the mixed staffing model, and data from evidence-based research are needed to attain stakeholder support for the model. To provide a reference for policymakers, further research will be necessary to clarify the impact of the mixed staffing model on hospital performance, patient outcomes, and nurse workloads.