The relationship between [doctors and nurses] is a major determinant of the quality of the healthcare practice environment. Both the safety and the quality of care patients receive depend upon the quality of the healthcare practice environment where care is provided (Joint Commission, 2008; Shen, Chiu, Lee, Hu, & Chang, 2010). Doctors and nurses make up the largest groups of individuals within this environment. The relationship between these two groups is a major determinant of the quality of the healthcare practice environment (Krogstad, Hofoss, & Hjortdahl, 2004; Joint Commission, 2008; Shen, et al., 2010). Abuse (verbal or physical) and intimidating or disrespectful behavior by doctors toward nurses impacts the healthcare practice environment in a negative way, affecting both nurse retention and patient outcomes (Institute of Medicine [IOM], 2010; Joint Commission, 2008; Schmalenberg & Kramer, 2009). It is important that healthcare organizations not only have a method in place to identify intimidating or disrespectful behaviors and a process for disciplinary action for offenders, but also a way to monitor relationships between nurses and physicians over time to determine the impact of any remediation interventions.

An electronic search of literature from the past decade, using CINAHL and Medline, identified studies about nurse and physician communication and/or collaboration and its impact on nurse retention and patient outcomes (Joint Commission, 2008; Manojlovich, 2005; Manojlocih & DeCicco, 2007; Rosenstein, 2002; Rosenstein, Russell, & Lauve, 2002). Studies that assessed relationships between nurses and physicians did so by examining the presence or absence of abusive and intimidating events by physicians, as perceived by nurses (Rosenstein, 2002; Rosenstein & O’Daniel, 2005). No studies were found that looked at the positive aspects of the healthcare practice environment associated with nurse and physician relationships; and no studies assessed the relationships between nurses and physicians in the same healthcare practice environment as perceived by both nurses and physicians at the same time, and using the same instrument.

Our research sought to address this gap using a large, two-group-comparative study. Our goals were (a) to explore nurses’ and physicians’ perceptions of their professional relationships within the healthcare practice environment in which they were employed; (b) to determine if nurses and physicians had different perceptions of behaviors associated with respect; and (c) to discover if relationships between nurses and physicians impacted nursing practice decisions.

The research questions were:

1. Do nurses and physicians differ in their assessment of the quality of the healthcare practice environment where they are employed?
2. Do nurses and physicians differ in their perceptions of respectful behavior?
3. Do perceptions of nurse and physician relationships affect practice decisions made by nurses?

Review of the Literature

The professional practice environment has been implicated as a factor that affects nurse retention and recruitment (Galletta, Portoghese, Battistelli, & Leiter, 2013; Joint Commisssion, 2008), as well as patient outcomes (Friese & Manojlovich, 2012; IOM, 2010; Joint Commission, 2008; Manojlovich & DeCicco, 2007; Rosenstein & Naylor, 2012; Schmalenberg & Kramer, 2009; Tschannen & Kalish, 2009). It has been posited that those organizations with a positive professional practice environment, characterized by healthy and respectful nurse-physician relationships, are better able to recruit and retain the best nurses; and that this, coupled with higher levels of communication, respect, and collaboration between nurses and physicians, contribute to a better environment for patients (Galletta et al., 2013; Nelson, King, & Brodine, 2008). Although previous studies have examined the impact of various aspects of the professional practice environment on nurse satisfaction, nurse retention, nurse recruitment, and patient outcomes (Nelson, et al., 2008; Rosenstein & Naylor, 2012; Manojlovich & DeCicco, 2007), our study sought both to assess nurse and physician perceptions of the presence of positive characteristics of nurse-physician relationships and to identify factors that impact, or are impacted by, nurse-physician relationships.

The professional practice environment is affected by the historical development of the nursing and medical professions and societal norms... The professional practice environment (PPE) model (Figure) proposed by Siedlecki and Hixson (2011) was used as the theoretical base for this study. According to this model, the professional practice environment is the place where nursing and medical care take place, and perceptions of relationships between nurses and physicians is a good indicator of the quality of the practice environment. The professional practice environment is affected by the historical development of the nursing and medical professions and societal norms; thus time and geographical location impact the professional practice environment and the people who practice within it.

Figure. Positive Professional Practice Environment Scale© and Four Subscales

The PPE model (Siedlecki & Hixson, 2011) is supported by the assumption that a positive professional practice environment is an environment that is more than and different from the absence of disruptive, rude, disrespectful and abusive behavior, and one that impacts nurse and physician satisfaction and patient outcomes. According to this model the quality of the professional practice environment can be assessed by examining nurses’ and physicians’ perceptions of patterns of mutual respect, communication, and collaboration between nurses and physicians within the practice environment.

The Study

In this section, we will present the measures we used to assess perceptions of the quality of the healthcare environment and the steps we took to protect our human subjects. We will also describe our research and data analysis procedures, along with assumptions made in this study.

Assessment Measures

The Professional Practice Environment Assessment Scale (PPEAS) was used to assess nurse and physician perceptions of the quality of the healthcare environment (Table 1). The PPEAS (Siedlecki & Hixson, 2011) is a 13-item scale developed to assess the quality of the professional practice environment at a single point in time. It looks at the presence of positive physician and nurse characteristics, organizational characteristics (beliefs about the importance of nurse-physician respect, communication, and collaboration on patient outcomes), and frequency of joint-patient-care decision making.

A full description of the development and psychometric testing of the PPEAS has been previously reported (Siedlecki & Hixson, 2011). The 13 items in the PPEAS are worded so it does not matter if the respondent is a nurse or physician; respondents are asked to rate their agreement with each item using a scale of 1 to 10. Larger numbers indicate a more positive perception of the presence of that element in the environment. The overall quality of the professional practice environment is assessed by summing the 13 items. Scores can range from 13 to 130, with higher scores indicating a more positive professional practice environment. An optimal (highest possible score) professional practice environment is indicated by a score of 130 (100%). Scores are standardized (0 to 100) by converting the raw score to a percentage to allow for easier comparisons.

Construct validity of the PPEAS was previously established through principle component analysis with varimax rotation that confirmed a four-factor solution explaining 72% of the variance (Siedlecki & Hixson, 2011). Internal consistency of the PPEAS in this study was confirmed using Cronbach's alpha, determined to be 0.856 for the nurse group and 0.842 for the physician group. This suggests it was a reliable measure in this sample.

The PPEAS examines perceptions of evidence of mutual respect experienced in the professional practice environment; however it was unclear if nurses and physicians would differ in their beliefs about what respectful behavior looks like. To determine what behaviors nurses and physicians considered respectful, we asked a single, forced-choice question with six possible responses. The responses were developed from a review of the literature, input from nurses and physicians, and consensus using a two-stage Delphi technique (CVI = .92).

Finally, to determine if behaviors and attitudes of individual physicians might impact nursing practice decisions, we posed a single question to nurse respondents. This single forced-choice (yes/no) item asked nurses whether “a physician’s behavior or attitude affects how they ask a question or if it makes them reluctant to report changes in a patient’s clinical condition.”

Protection of Human Subjects

Institutional review board approval was obtained to conduct an anonymous electronic survey. Consent was implied if respondents submitted a survey. No identifying information was included on the survey form; even the researchers were not aware of the identity of individual respondents.

Procedures

A convenience sample of all inpatient and outpatient nurses and physicians at a large Midwestern, Magnet® Designated, healthcare facility were identified, and invitations to participate were distributed via in-house email. It is estimated that this email invitation was sent to 4,130 nurses and physicians. Because nurses and physicians receive dozens of emails each day, they often do not even open the ones that do not appear to be important. Hence the email invitations were distributed to each group (nurses and physicians) by their respective directors, to maximize likelihood that the email invitation would be considered important and would be opened. Data collection lasted 30 days, with weekly reminders sent out for the four weeks of the study.

Responses were automatically entered into a database by the survey software (Zoomerang©) and downloaded for analysis. Incomplete surveys, which were excluded from analysis, resulted in an attrition rate of three percent (n = 41), and were evenly distributed between nurses (n = 21) and physicians (n = 20). Data was exported into an SPSS (Version 20) database for analysis.

Data Analysis

The sample was described by measures of central tendencies and frequencies. Range, mean, and standard deviation were used to describe the professional practice environment; frequencies were used to identify behaviors that most demonstrated respect and to determine the extent to which physician behavior or attitude impacts nursing practice decisions. Analysis of variance (ANOVA) and chi-square analysis were used to compare nurse and physician groups on major study variables. Multivariate analysis of variance (MANOVA), using group (nurse-physician) as the independent variable and the four PPEAS subscales as dependent variables, were used to assess observed differences between nurses and physicians. Because of the large sample size, a p-value <0.01 was selected to determine statistical significance.

Prior to data analysis, ANOVA and MANOVA assumptions were assessed. Independence of observations was controlled by study design. The assumption that the dependent variables are normally distributed was confirmed with visual inspection of histograms and Q to Q plots. The assumption of linearity was confirmed using visual inspection of bivariate scatterplots; and the assumption of homoscedasticity was confirmed with visual examination of box plots (Mertler & Vannatta, 2010). Finally, to assure that there was no multicollinearity, correlations were examined. In this sample, we found a significant correlation (p <.05) between all variables (groups) but none were greater than r = .30. Correlations greater than r = .30 are an indication of multicollinearity (Mertler & Vannatta, 2010).

Findings

In this section, we will describe the characteristics of our sample. We will also compare perceptions of the quality of the healthcare practice environment, describe differences between the groups, report behaviors that best demonstrate respect, and discuss the impact of physician behaviors on nurse behaviors.

Sample Characteristics

The response rate for this survey was 33%; resulting in a total sample of N = 1,364. The final sample was composed of 822 nurse and 542 physician responses (Table 2). All respondents were employees of a large healthcare center in the Midwest. Most nurses were female (n = 732; 95%) and most physicians were male (n = 372; 72%); this difference was significant (p <.001). The mean age for the sample (nurses and physicians) was 43.91 (SD = 10.87) and both groups (nurses and physicians) had many years of professional experience (x = 17.96; SD = 10.87). Although the groups did not differ by age (p = 0.156), they did differ by years of professional experience (p <0.001) with nurses having more experience (x = 18.99; SD = 10.74) than physicians (x = 16.42; SD = 10.88). The majority of nurses were staff nurses (n = 479; 62%) and the majority of physicians were staff physicians (n = 411; 76%). Only 4% of nurse respondents were LPNs, the remainder were RNs, with 68% (n = 547) having at least a bachelor’s degree.

Table 2. Demographic Characteristics

CNS = Clinical Nurse Specialist; NP = Nurse Practitioner; LPN = Licensed Practical Nurse;
AD = Associate Degree; M = mean; SD = Standard deviation

Comparison of Perceptions of the Quality of the Professional Practice Environment

...nurses were more likely to hold stronger views on the impact of respect, communication, and collaboration on patient outcomes (organizational characteristics), than were physicians. Multiple analysis of variance (MANOVA) revealed a significant (Wilk’s Lambda p < 0.001) overall difference between groups (physicians and nurses) for the combined dependent variable (four subscales combined). Follow-up ANOVA with a Bonferroni adjustment found a significant difference between groups for all subscales (Table 3). In general, physicians rated the overall environment better than nurses, and they scored physician characteristics and patient care decision making better than nurses. However, nurses were more likely to hold stronger views on the impact of respect, communication, and collaboration on patient outcomes (organizational characteristics), than were physicians. To allow for easier comparisons between groups the raw scores were converted to percentages, resulting in a standardized score for each subscale and for the total scale. The mean standardized score for nurses was 73%, while the mean standardized score for physicians was 77% and this difference was significant (p <.001).

Table 3. Comparison of Physicians and Nurses Perceptions of the Professional Practice Environment

[View full size]

Within Group Comparisons

Using split file analysis techniques, ANOVA was used to assess the effects of nursing education, nursing role, and physician role on overall perceptions of the quality of the professional practice environment. Higher nursing education was not significant (p = 0.153); but nursing role was found to be significant (p = 0.017), with staff nurses and nurse managers rating the quality of the professional practice environment lower than clinical nurse specialists and nurse practitioners (Table 4). Examining individual subscales, only perceptions of physician characteristics differed significantly by nursing role, with nurse practitioners having more favorable perceptions of physician characteristics than nurses in other roles (76%, p = 0.008).

Table 4. Nurse Perceptions of Professional Practice Environment

[View full size]

Physician role was also found to be significant (p <0.001), with staff physicians rating the professional practice environment more favorably than residents or fellows (Table 5). When examining individual subscales, no significant differences were noted in perceptions of physician characteristics or organizational characteristics by role. However, staff physicians rated nurse characteristics (77%, p <0.001) and presence of positive characteristics of patient care decision making (80%, <0.001) higher than did residents or fellows.

Table 5. Physician Perceptions of Professional Practice Environment

M = Mean; SD = Standard deviation

Behaviors that Best Demonstrate Respect

Descriptive statistics and chi square analysis (Table 6) were used to examine the behaviors nurses and physicians indicated best demonstrated respect. Both physicians and nurses identified “treats me with civility, courtesy, and respect” most often (54% by both groups). Nurses’ second most frequently cited behavior was “listens to me when I am speaking” (17%), while the second choice by physicians was “works to solve problems without hostility or verbal abuse” (26%).

Table 6. Behaviors that Best Demonstrate Respect

Impact of Physician Behavior on Nurses’ Behavior

...staff nurses [were] most likely to be affected by a physician’s behavior and the advance practice nurse the least likely. Lastly, nurses were asked whether “a physician’s attitude affects how you ask a question or makes you reluctant to report a change in a patient's clinical condition?” to examine what effect physician behavior might have on nursing care behaviors. Of the 807 nurses who answered the question, 55% (n = 479) responded ‘yes’. Further analysis of this question found that level of education of the nurse was not predictive of a yes or no response (p = 0.934); however, nursing role (p <.001) was predictive, with staff nurses most likely to be affected by a physician’s behavior and the advance practice nurse the least likely. Exploratory analysis also found younger nurses (<45 years of age) and/or less experienced nurses (<20 years of experience) were more likely to report being affected by negative physician behaviors than older or more experienced nurses (p = .001).

Strengths and Limitations

The strength of this study was the large sample size and the inclusion of both nurses and physicians. Limitations relate primarily to generalizability. This study was undertaken at a large Magnet® Designated healthcare facility where physicians are salaried employees, potentially limiting generalizability to other similar facilities. The return rate is another limitation, as those who chose to respond likely had stronger perceptions (positive or negative) than those who did not respond.

Discussion

This study improves upon previous work by examining the presence of positive professional practice environment characteristics, rather than simply the absence of negative, rude, or disrespectful behaviors. In addition, in this study we compared nurses’ and physicians’ perceptions of the same environment, at the same time, and using the same scale. Additionally, we sought to link perceptions of the quality of the practice environment to patient care outcomes through exploration of nursing behaviors.

...nurses and physicians perceive their work environment differently, and neither group perceives their practice environment as optimal... We found nurses and physicians perceive their work environment differently, and neither group perceives their practice environment as optimal (100%), suggesting significant room for improvement. This is consistent with findings reported by others (Friese & Manojlovich, 2012; Joint Commission, 2008; Krogstad et al., 2004). The professional practice environment assessment data was normally distributed in this study for physicians and nurses, suggesting that while many perceive their environment positively, a similar number perceived their environment as less positive.

Comparisons between our study and previous studies are difficult due to differences in sampling, sample size, measurement, and analysis. Most studies surveyed only nurses, and small sample sizes hampered their ability to find differences. In addition, previous studies used instruments designed to measure the absence or presence of negative behaviors, primarily by physicians. Our study sought to improve upon previous work by exploring this phenomenon from both physicians’ and nurses’ perspectives simultaneously.

Our data suggested that physician behaviors and attitudes may directly affect nursing patient care behaviors, and consequently, patient outcomes. This finding is in agreement with the findings reported in the Joint Commission Sentinel Event Alert (Joint Commission, 2008). It is also consistent with findings reported by Rosenstein and O’Daniel (2005) and Rosenstein and Naylor (2012).

Acknowledging differences in values, incentives, and perceptions can provide insights to focus improvement initiatives. Although the finding that nurses and physicians differed in their perceptions of the quality of the professional practice environment is not surprising, it is instructive. The professional practice environment is a shared work setting. Acknowledging differences in values, incentives, and perceptions can provide insights to focus improvement initiatives. Accommodating these different perspectives can improve the likelihood of success by not framing the initiative in a way that views the other group as the problem to be solved, but rather as a full partner in devising sustainable solutions.

Conclusion

Without establishing new and desirable behavioral norms, old habits and behaviors are likely to fill the vacuum temporarily left by an ‘eliminate the negative’ initiative. Initiatives with the aim of improving the nurse-physician relationships by eliminating rude and disrespectful behavior are worthwhile and can produce demonstrable improvements (Joint Commission, 2008). However, we suggest they are incomplete without at least an equal emphasis on promoting new behaviors that encourage a positive professional practice environment and healthy, respectful nurse-physician relationships. Without establishing new and desirable behavioral norms, old habits and behaviors are likely to fill the vacuum temporarily left by an ‘eliminate the negative’ initiative. Consequently, simply assessing the environment by the frequency of poor behavior is of little help to inform the design of programs to improve and enhance the positive nurse-physician relationships. 

Authors

Sandra L. Siedlecki, PhD, RN, CNS
Email: siedles@ccf.org

Sandra L. Siedlecki is a Senior Nurse Scientist in the Office of Nursing Research and Innovation at the Cleveland Clinic. She has over 40 years of experience as a clinical nurse, an educator, and a clinical researcher. Dr. Siedlecki has worked with multidisciplinary teams to investigate best practices as it relates to professionalism and professional behavior within a healthcare setting. Dr. Siedlecki received her BSN and MSN from the University of Akron, and her PhD in Nursing from Frances Payne Bolton School of Nursing at Case Western Reserve University in Cleveland, Ohio.

Eric D. Hixson, PhD, MBA
Email: Hixsone@ccf.org

Eric Hixson is the Senior Program Administrator in Business Intelligence at the Cleveland Clinic and has 18 years of experience in developing and implementing operational and clinical quality reporting initiatives, data warehousing, registry implementation, and operations. Currently, Eric is responsible for developing predictive analytics strategy and incorporating metrics into the current BI infrastructure for strategic initiatives and operational reporting. He develops and implements predictive models, forecasting, and data mining algorithms to support enterprise priorities. He has led development and continued expansion of operational reporting that leverages the electronic medical record content with business intelligence capability. The success of Cleveland Clinic’s Enterprise Business Intelligence is featured as a case study by Howard Dresner in Profiles in Performance: Business Intelligence Journeys and the Roadmap to Change. Eric received his PhD from Case Western Reserve University in Epidemiology and Biostatistics and MBA from Cleveland State University in Healthcare Administration.