Introduction

Most people in Taiwan (99.93%) are enrolled in the National Health Insurance single-payer system (National Health Insurance Administration, Ministry of Health and Welfare, Taiwan, ROC, 2020), which is designed to make access to healthcare easy and inexpensive for the population. In addition, Taiwan is facing a rapidly growing aging population (20% of the population in 2026 is expected to be over 65 years old), increased acuity in hospitalized patients, working hour restrictions for residents and physicians, and consumer demand for care that meets minimum quality and safety standards, all of which have led to increasing demand for healthcare services. Addressing the growing healthcare strains, the Taiwan Government has authorized formal nurse practitioner (NP) regulations and programs. NPs have provided healthcare to inpatients via doctor of medicine–NP (MD-NP) team-based practice since 2006 (Chiu et al., 2016; Tsay et al., 2019). NPs provide nursing and medical care to meet the complex needs of patients and their families using a holistic, health-centered approach. NPs represent an important part of the workforce and work with physicians to deliver team-based care. They contribute to meeting the growing demand for healthcare in acute care practices. However, there is controversy over the expanding role of NPs in acute care, as this trend may challenge the current teamwork relationship between NPs and physicians. To date, little evidence is available regarding the best way to facilitate teamwork between NPs and physicians in acute care hospital settings.

The team-based care or practice model was implemented in response to changes in the healthcare system, including an increasing number of older patients with chronic diseases and the multidecade rise in medical costs per patient (Mitchell et al., 2012). However, team-based care, multidisciplinary care, collaborative care, and multispecialty care have often been used interchangeably in the literature (van der Marck & Bloem, 2014). The Institute of Medicine defined team-based care as “the provision of health services to individuals, families, and/or communities by at least two healthcare providers who work collaboratively with patients and their caregivers to accomplish shared goals within and across settings to achieve coordinated high-quality care” (as cited in Mitchell et al., 2012, p. 5). Baik (2017), utilizing a concept analysis framework, revealed that team-based care embraces interprofessional collaboration, a patient-centered approach, and integrated care processes. This is accomplished by understanding the roles and responsibilities of other team members, mutual respect, and organizational support. Team-based practice is vital to providing patient-centered care and to promoting care quality, safety, and improved patient experiences (Mitchell et al., 2012). In addition, a growing body of research indicates that team-based practice is associated with improved patient outcomes such as decreased hospital readmission rates for high-risk patients and reduced length of hospital stays (Brandt et al., 2014; Mitchell et al., 2012).

NPs frequently work with physicians in team-based practices in hospitals (Cowan et al., 2006). However, NP practice autonomy is often misinterpreted and is a source of confusion that impedes professional development (American Association of Nurse Practitioners, 2018). Recently, Peacock and Hernandez (2020) conducted a concept analysis that redefined practice autonomy as NPs' use of their experience, clinical judgment, and responsibility to practice without restriction in professional collaboration with other healthcare professionals. Autonomy allows NPs to practice to the maximum extent of their advanced education and to continue to influence healthcare outcomes. The American Association of Nurse Practitioners categorized NP practice into full practice, reduced practice, and restricted practice (American Association of Nurse Practitioners, 2018). Full practice allows NPs to evaluate patients; to diagnose, order, and interpret diagnostic tests; and to prescribe medications. Reduced practice requires a regulated, collaborative agreement to provide patient care. Restricted practice requires supervision, delegation, or team management to provide patient care. Although NPs have the potential to fill the gap in care in Taiwan, their practice autonomy continues to be threatened in large part by the Physician Practice Act, which restricts nonphysicians from providing medical care to patients. In other words, the practice of NPs in Taiwan continues to fall into the “restricted practice” category based on its team-based care requirements.

The effect of NP autonomy may be positive or negative. The potential, positive effects include eliminating patient barriers, improved cost-effectiveness of care, increased NP job satisfaction, increased quality of patient care, and improved efficiency of MD-NP team-based practice. The potential, negative effects include decreased NP job satisfaction, lower empathy, and professional burnout (Peacock & Hernandez, 2020; Poghosyan et al., 2014). The autonomy level of NPs differs between primary care and acute care settings. A national survey of NP autonomy (N = 8,311) found that NPs working in primary care reported the highest levels of autonomy, whereas those working in hospital settings reported the lowest levels (Athey et al., 2016). In contrast, a small study of acute care NPs (n = 54) in different specialty areas in a large metropolitan hospital reported that NPs have high, overall levels of autonomy (Cajulis & Fitzpatrick, 2007). Nonetheless, the level of autonomy of NPs in acute care settings in Taiwan is an issue that has not yet been explored in the literature. In addition, autonomy, specifically feeling that one's NP skills are fully utilized, was found to be an important predictor of efficacy in team-based practice (Athey et al., 2016). Little is known about the factors affecting the team-based practice of NPs. Examining these relationships is expected to provide insights into the strength of NP practice.

Leadership has a potentially vital role to play in promoting team-based practice by engaging members in collaborative efforts to better meet patient needs. One of the best known forms of leadership, transformational leadership, has been shown to be effective in maximizing the performance and outcomes of individuals and team-based practice (Fischer, 2016). NPs achieve leadership by empowering others and highlighting the importance of interpersonal relationships within the care team. In practice, leadership approaches often involve setting a vision and motivating people to achieve common goals and focus on working collaboratively. NPs see themselves as group members rather than exerting hierarchical power over others in practice. Poghosyan and Bernhardt (2018) investigated transformational leadership in 278 primary care NPs, finding that nearly 50% reported that leaders did not share information equally between NPs and physicians. They also found that 46% reported that NPs were not represented on important organizational committees. Perceptions of team-based practice and leadership are mutually related and may impact the ability of clinicians to provide patient care safely (Manser, 2009). Researchers have recommended that leaders in primary care practices consider applying transformational leadership principles to promote NP practice. Van Kraaij et al. (2020) explored the perceptions of the leadership roles of NPs in Dutch hospital care, finding that NP leadership mainly related to direct clinical patient care and that leadership at the professional and organizational levels appeared to be underutilized. However, the leadership of NPs needs further exploration in Taiwan.

The relationships among the autonomy, leadership, and team-based practice of NPs have been infrequently studied, particularly in acute care settings. In one recent study, 163 primary care practices in Massachusetts were surveyed to investigate whether the autonomy of NPs and their relationships with leadership affect the relationship between NPs and physicians within team-based practice settings (Poghosyan & Liu, 2016). They found that autonomy and leadership significantly predicted teamwork between NPs and physicians, explaining 41% of the variance in team-based care. Nonetheless, they collected data in one state only, which limits the generalizability of their results to NPs in other primary care settings. On the basis of the above, a gap exists in the research with regard to the actual level of NP involvement in team-based practice. Researchers need to address that gap by exploring the complex factors that influence team-based practice. This study was designed to examine factors that predict the efficacy of NP team-based practice in acute care settings using a large, national sample of NPs. Specifically, we were interested in examining the extent to which autonomy and leadership relate to team-based practice. Elucidating the relationships among professional autonomy, leadership style, and team-based practice is important to better supporting clinical practice, as NPs continue to struggle to engage in the full scope of practice in Taiwan.

Methods Study Design, Setting, and Participants

This study used a cross-sectional design and a national survey to investigate and explore the relationships among autonomy, leadership style, and team-based practice in acute care NPs in Taiwan. Data were collected using an anonymous structured questionnaire provided as an online survey from April 1 to May 30, 2020.

The participants were drawn from the nationwide database of the Taiwan Association of Nurse Practitioners (TANP), which includes approximately 95% of NPs in Taiwan. Eligibility criteria for participants were as follows: (a) active member of TANP, (b) national NP certified, and (c) working in clinical practice for at least 1 year as an NP in an acute care setting. The 6,000 eligible, active members of TANP all received invitations to participate via email. NPs who agreed to participate in this study signed an informed consent form before taking the online survey. They completed the questionnaires on the online survey platform SurveyCake (25sprout, Taipei City, Taiwan). The questionnaires included demographic and practice characteristics, the Dempster Practice Behavior Scale (DPBS), the Multifactor Leadership Questionnaire (MLQ-6S), and the NP–physician relations (NP-PR) subscale of the Nurse Practitioner Primary Care Organizational Climate Questionnaire (NP-PCOCQ). One thousand four hundred forty-two NPs working at 181 hospitals completed the survey (response rate: 24.03%). The 1,391 participants who completed the survey without missing data were included in the analysis. Considering the statistical power, the minimal sample size estimated using G*Power 3.1 under the setting (power = .90, α = .05, effect size [f2] = .10, 10 predictors) for multiple regression analysis was 215. We estimated about a 20% no-response rate. Therefore, at least 258 participants were necessary, and the sample size acquired (n = 1,391) greatly exceeded this number. The institutional review board of the hospital approved this study (approval number: 2020012).

Measures Demographic characteristics

Demographic data collected included age, years of experience, annual salary (New Taiwan dollar [NTD]), patient load (per day), overtime hours, number of NPs in hospital, gender, education, marital status, care model, shift work status, intention to leave in the coming 1-year period, background of practice managers, hospital level, and work region. For the hierarchical linear model (HLM) analysis, the organization-level and individual-level effects on team-based practice were investigated.

Autonomy

The DPBS, which was designed to measure the extent to which autonomous behaviors occur in practice, was used to measure autonomy in the NPs (Dempster, 1990). This scale comprises 30 items rated on a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). The scale comprises the four subscales of readiness, empowerment, actualization, and valuation. Readiness measures competence, skill, and mastery; empowerment measures rights and privileges; actualization measures decision making, accepting responsibility, and accountability for actions; and valuation measures self-respect, value, worth, achievement, and satisfaction. The DPBS total possible score ranges from 30 to 150, with higher scores corresponding with higher levels of autonomy. The DPBS has shown adequate reliability and validity in previous research (Dempster, 1990, 1994; Goolsby et al., 2020). The Cronbach's α was .93 for the DPBS in this study.

Leadership

The MLQ-6S is widely used to measure leadership style in clinical practice (Bass & Avolio, 1992, 2004). This questionnaire comprises 21 items rated on a 5-point Likert-type scale ranging from 1 (not at all) to 5 (frequently if not always). MLQ-6S has seven subscales that measure three different leadership styles. The transformational leadership scale comprises the factors of idealized influence, individualized consideration, intellectual stimulation, and inspirational motivation, whereas the transactional leadership scale consists of contingent reward and management by exception. A further scale deals with laissez-faire leadership as passive/avoidant leadership.

The score of each subscale (three items) may be categorized as high (12–15), moderate (8–11), or low (1–7), with higher scores corresponding to higher levels of leadership style, as perceived by NPs in practice. Satisfactory reliability and validity have been reported for the MLQ-6S in previous studies (Bass & Avolio, 2004; Moon et al., 2019; Tejeda et al., 2001). The Cronbach's α was .91 for this questionnaire in this study.

Team-based practice

Team-based practice was evaluated in this study using the NP-PR subscale of the NP-PCOCQ. The NP-PCOCQ was developed based on evidence-based practice and interviews with NPs (Poghosyan et al., 2013). The relationship between NPs and physicians is critical in care settings because of the close collaboration required among these healthcare providers (Poghosyan et al., 2013). NP-PR is able to access information on the relationship, communication, and teamwork between NPs and physicians, which may act as team-based practice in the work setting. The NP-PR subscale consists of seven items rated on a 4-point Likert-type scale from 1 (strongly disagree) to 4 (strongly agree), with higher scores indicating higher levels of team-based practice. This subscale has been shown to provide satisfactory reliability (Cronbach's α: .87–.95) and validity (Poghosyan et al., 2013, 2017). The Cronbach's α of the NP-PR was .91 in this study.

Data Analysis

The descriptive statistics in this study were expressed in terms of percentage, mean, and standard deviation (SD), and Pearson product–moment correlations were performed among the continuous variables. An independent-samples t test was computed to examine the differences in continuous variables between the two groups.

To control for the potential influence of individual acute-care hospital settings on team-based practice (i.e., the organizational effect), the HLM was used to separately assess the organizational and individual (NP) effects on team-based practice. The 1,391 NP participants (Level 1) were nested within the 181 acute care hospitals (Level 2) in line with the recommendations of Hox (1998). First, the null model was generated to detect the presence of an organizational effect on the practice level of team-based practice. Next, a chi-square test was applied on the Level 2 residual variance and the intraclass correlation coefficient (ICC) indexes, including ICC(1) and ICC(2), to examine the rationale of the HLM analysis. ICC(1) and ICC(2) values of > .059 and > .6, respectively, are required to conduct HLM analysis (Glick, 1985; Hofmann, 1997). If the ICC indexes of the null model met the criteria, the HLM analysis was performed. If the organization level was shown to not significantly affect teamwork, multiple regression analysis was deemed appropriate for examining the association between the individual characteristics of NPs and team-based practice.

All of the variables found to be significantly associated with team-based practice were included in the stepwise multiple regression analysis. The statistics of the regression analysis, including regression coefficients (B), standardized regression coefficients (β), standard error (SE), ΔR2, and R2, were estimated. The p values Values (two-sided tests) less than .05 were statistically significant. Statistical analyses were performed using IBM SPSS Statistics Version 22.0 (IBM Inc., Armonk, NY, USA) and HLM Version 7 (Scientific Software International, Skokie, IL, USA).

Results

The characteristics of the participants and the associations between these characteristics and team-based practice are presented in Table 1. The mean age of the participants was 41.72 years (SD = 5.37; range: 26–58 years), mean NP experience was 7.58 years (SD = 4.41; range: 1–15 years), mean annual salary was 798,000 NTD (SD = 15.55; range: 42–150 NTD), and mean day-shift patient load (number per day) was 12.71 (SD = 9.11; range: 1–51). The hospitals covered in this study employed an average of 81.04 NPs (SD = 59.75; range: 1–350), most of whom were female (95.8%), married (66.1%), and involved in MD and NP team-based practice. Significantly over one in 10 participants (15.6%) indicated that they intended to leave the nursing profession within 1 year. Many of the participants (67.5%) directly reported to managers with no NP background, and many practiced outside the medical center (62.3%) and northern Taiwan (57.3%).

Table 1 - NP- and Organization-Level Characteristics and Association With Team-Based Practice (N = 1,391) Variable n % Team-Based Practice p Mean SD t/r Age (year) 41.72 5.73 .10 < .01** Years as NP 7.85 4.41 .14 < .01** Annual salary (10,000 NTD) 79.80 15.55 .13 < .01** Patient load (per day shift) 12.71 9.11 −.05 .07 Overtime hours 0.96 0.69 .02 .55 Number of NPs in hospital 81.04 59.75 .08 .02* Gender  Female 1,333 95.8 22.04 2.99 −1.22 .23  Male 58 4.2 21.57 2.89 Graduate degree  Yes 262 18.8 22.75 3.19 4.22 < .001***  No 1,129 81.2 21.85 2.79 Marital status  Married 919 66.1 22.05 2.81 0.48 .63  Unmarried 472 33.9 21.97 3.03 Care model  VS + R + NP 467 33.6 22.29 2.90 −2.45 .01*  VS + NP 924 66.4 21.89 2.88 Shift work  Yes 611 43.9 22.03 2.89 0.16 .87  No 780 56.1 22.01 2.89 Intention to leave within 1 year  Yes 217 15.6 20.91 3.05 −5.89 < .001***  No 1,174 84.4 22.22 2.81 Managers  With NP background 452 32.5 22.13 2.86 0.97 .33  Without NP background 939 67.5 21.97 2.90 Hospital level  Medical center 525 37.7 22.30 2.88 2.83 .01*  Nonmedical center 866 62.3 21.85 2.88 Geographic region  Northern 594 42.7 22.01 2.86 −1.12 .91  Other 797 57.3 22.03 2.92

Note. NTD = New Taiwan dollar; VS = visiting staff; R = resident; NP = nurse practitioner.

*p < .05. **p < .01. ***p < .001.

Participant age (r = .10, p < .01), years of NP experience (r = .14, p < .01), annual salary (r = .13, p < .01), number of NP employees working at the hospital (r = .08, p = .02), educational level (t = 4.22, p < .001), care model used (t = −2.45, p = .01), intention to leave within 1 year (t = −5.89, p < .001), and hospital level (t = 2.83, p = .01) were all significantly associated with team-based practice.

Descriptive statistics for all of the scales are shown in Table 2. The means of the autonomy dimensions indicate that the level of autonomy among the participants was lowest in empowerment (item mean = 2.45, SD = 0.49), indicating that they perceived low levels of empowerment in practice. The item mean scores on the leadership subscales were highest for idealized motivation (item mean = 3.91, SD = 0.53), followed by laissez-faire leadership (item mean = 3.32, SD = 0.58). The item mean score for team-based practice was 3.15 (SD = 0.45), indicating a moderately high level of team-based practice. All of the subscales for autonomy (ranging from .39 to .61, p < .001) and leadership (ranging from .09 to .41, p < .001) showed significantly positive correlations with team-based practice.

Table 2 - Descriptive Statistics: Autonomy, Leadership, and Team-Based Practice Variable Subscale Score Items Correlation With Team-Based Practice Mean SD Mean SD r Autonomya  Readiness 41.85 5.07 3.80 .46 .55***  Empowerment 17.17 3.42 2.45 .49 .39***  Actualization 36.10 3.83 4.01 .43 .55***  Valuation 11.66 1.67 3.89 .56 .53***  Total 106.78 11.69 – – .61*** Leadershipb  Idealized influence 11.74 1.58 3.91 .53 .41***  Inspirational motivation 10.88 1.79 3.63 .60 .38***  Individual stimulation 10.39 1.96 3.46 .65 .28***  Individual consideration 10.15 1.89 3.38 .63 .25***  Contingent reward 10.35 2.02 3.45 .67 .26***  Management-by-exception 10.99 1.56 3.66 .52 .28***  Laissez-faire leadership 9.95 1.74 3.32 .58 .09**  Total 74.47 9.28 – – .37*** Team-based practice 22.02 2.89 3.15 .45 –

a Item mean in autonomy ranged from 1 to 5 (1 = strongly disagree, 2 = disagree, 3 = undecided, 4 = agree, and 5 = strongly agree). b Item mean in leadership ranged from 1 to 5 (1 = never, 2 = rarely, 3 = sometimes, 4 = often, and 5 = always).

**p < .01. ***p < .001.

The results of the HLM analysis are shown in Table 3. The random effect (τ00) reached significance (χ2 = 258.31, p < .001). However, the ICC(1) showed a low correlation among the different organizations (ICC[1] = .06), and the ICC(2) was also far less than .6 (ICC[2] = .26), which does not support the use of multilevel modeling. Therefore, organization-level effects may have a minimal impact on the team-based practice of NPs. Multiple regression analysis was thus more appropriate for examining the association between NP characteristics and team-based practice (Table 3).

Table 3 - The Results of Null Model in NPs' Team-Based Practice Estimation Coefficient SE t Ratio p Fixed components 211.91 < .001***  Intercept, β0 – –  Intercept 2, γ00 22.00 0.10 Variance of random components χ2 = 258.31 < .001***  u0j(τ00) 0.51 0.72  Residual 7.77 2.79 ICC(1) 0.06 ICC(2) 0.26

Note. Level 1: teamwork ij = β0j + rij.; Level 2: β0j = γ00 + u0j.; mixed model: teamwork ij = γ00 + u0j + rij. ICC(1) = intraclass correlation coefficient (1), assesses the possibility of HLM analysis; ICC(2) = intraclass correlation coefficient (2), assesses the reliability of group-level means; NPs = nurse practitioners.

***p < .001.

The results of the stepwise multiple linear regression indicated that the participants with greater autonomy in actualization (β = .26, p < .001), empowerment (β = .20, p < .001), and readiness (β = .20, p < .001) had better team-based practice. In addition, the idealized influence leadership style (β = .13, p < .001) and hospital level (β = .06, p < .001) were shown to improve team-based practice performance. The final model explained 39% of the variance in team-based practice (F = 8.93, p < .001, R2 = .39), whereas autonomy in actualization and empowerment were the two most important predictors of team-based practice, explaining about 35% of the variance (Table 4).

Table 4 - Results of Stepwise Multiple Linear Regression in the Team-Based Practice of NPs Step B SE R 2 ΔR 2 F β t