| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Corresponding author: Dr Milisa Manojlovich, University of Michigan School of Nursing, 400 N Ingalls, Room 4306, Ann Arbor, MI 48109-0482 (e-mail: mmanojlo{at}umich.edu).
 |
Abstract |
|---|
 Top Abstract ObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
Objectives To examine the relationships between nurses’ perceptions of their practice environment, nurse-physician communication, and selected patients’ outcomes.
Methods A nonexperimental, descriptive design was used, and all nurses (N=866) working in 25 intensive care units in southeastern Michigan were surveyed. The Conditions for Work Effectiveness Questionnaire-II and the Practice Environment Scale of the Nursing Work Index were used to measure characteristics of the work environment; the ICU Nurse-Physician Questionnaire was used to measure nurse-physician communication. Nurses self-rated the frequency of ventilator-associated pneumonia, catheter-related sepsis, and medication errors in patients under their care.
Results A total of 462 nurses (53%) responded. According to multilevel modeling, both practice environment scales accounted for 47% of the variance in nurse-physician communication scores (P=.001). Nurse-physician communication was predictive of nurse-assessed medication errors only (R2=0.11). Neither environment scale was predictive of any of the patient outcomes.
Conclusions Healthy work environments are important for nurse-physician communication. In intensive care units, characteristics of the work environment did not vary enough to be significantly predictive of outcomes, suggesting that even in various types of critical care units, characteristics of the work environment may be more similar than different.
Certain elements in the hospital practice environment are important predictors of variation in patients’ outcomes.7 In nursing, hospital work environments have commonly been conceptualized from the perspective of the properties of magnet hospitals8 or of work-place empowerment.9 Extensive research from both perspectives has indicated a link between nurses’ working conditions and nurses’ and patients’ outcomes.10,11
Hospital nurses do not work in isolation; rather they collaborate and interact with other members of the healthcare team to provide quality patient care.12 Nurses and physicians make up the largest group of healthcare providers, and both daily confront complex problems with no easy solutions.13 However, communication between the professions does not flow as it should.14
In the classic study15 on outcomes of intensive care, communication between nurses and physicians was the single factor most significantly associated with excess hospital mortality. In more recent research,5 verbal miscommunication between nurses and physicians was responsible for 37% of all errors. Although other healthcare professionals are involved in quality patient initiatives, making the entire communication issue complex, nurses’ contribution to communication in the nurse-physician dyad is not fully understood. This study was framed from the nurses’ perspective. We gathered input solely from nurses and examined the relationship between nurse-physician communication and selected outcomes sensitive to nursing care.
The outcome data included nurses’ assessments of 2 nosocomial infections, ventilator-associated pneumonia (VAP) and catheter-associated sepsis, and nurses’ reports of medication errors. These outcomes already have been empirically linked to nursing processes.16 Nurses independently prevent infections by using hand washing and aseptic techniques as needed. Nurses must communicate with physicians when routine monitoring of a patient reveals a possible infection. Although the relationship between variables in staffing (ie, structure variables) and nosocomial infections has been examined,17,18 the relationship between variables in nursing processes, such as nurse-physician communication, and infections has not. In the ICU, rates of VAP can be as high as 86%,19 and catheter-associated sepsis may account for as many as 4000 ICU deaths annually.20 These statistics make a compelling case for investigating how nurse-physician communication can reduce the occurrence of both types of nosocomial infections.
| Communication between nurses and physicians is the one factor that is most associated with excess hospital mortality.
|
Administration of medications is primarily a role of nurses,21 and the second largest group of medication-related errors has been attributed to administration of drugs by nurses.22 Most medication errors are due to unsafe systems rather than to mistakes by individuals,23 but they tend to be underreported.22
The purpose of this study was to examine the relationships between ICU nurses’ perceptions of their practice environment, nurse-physician communication, and the outcomes of VAP, catheter-associated sepsis, and medication errors.
|
Objectives |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
|
Methods |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
| Most medication errors are due to unsafe systems rather than individuals’ mistakes.
|
Surveys were distributed during a 10-week period. Survey packets included a cover letter from the investigator, study instruments, a stamped return envelope, and, as a token of appreciation, a pocket reference guide from the American Association of Critical-Care Nurses for calculating medication doses. As an additional incentive to encourage participation, pizza parties were held in all nursing units in which at least 50% of the nurses participated in the survey. All answers were anonymous. Surveys were coded to identify the unit, but nurses’ names were not requested. The study was approved for protection of human subjects by the University of Michigan institutional review board and by the institutional review boards of the participating facilities.
Instruments
Workplace Empowerment.
Workplace empowerment refers to 4 social structures: opportunity, information, support, and resources embedded in any work environment that, when accessed, are sources of power. Workplace empowerment was measured by using the Conditions for Work Effectiveness Questionnaire-II. The questionnaire is a 6-subscale, 21-item measure of Kanter’s concept of empowerment.24,25 The tool uses a 5-point Likert-type scale. The 6 subscales are Opportunity, Information, Support, Resources, the Job Activities Scale II, and the Organizational Relationships Scale II. The first 4 subscales consist of 12 items (3 for each of Kanter’s 4 empowerment structures) and have high internal consistency.24 A total empowerment score is created by summing the scores for all 6 subscales (score range 6–30). Both content and construct validity of the Conditions for Work Effectiveness Questionnaire-II have been established. For this study, the Cronbach 
was .92.
Magnet Hospital Characteristics.
The extent to which hospitals’ nursing practice environments were consistent with the characteristics of magnet hospitals was measured by using the Practice Environment Scale of the Nursing Work Index.26 This tool uses a 4-point Likert-type scale. The Practice Environment Scale is based on the properties of magnet hospitals and consists of 5 subscales thought to address key domains in the hospital environment that support professional nursing practice: nurse participation in hospital affairs; nursing foundations for quality care; nurse manager ability, leadership, and support of nurses; staffing and resource adequacy; and collegial nurse-physician relations. Subscale internal consistency coefficients range from .71 to .84, with an overall Cronbach of .82.26 Construct validity has been established, and confirmatory factor analysis supports the 5-subscale structure of the tool. The National Quality Forum now recommends use of the Practice Environment Scale of the Nursing Work Index as a system-centered organizational effectiveness tool influenced by nurses’ care and performance. In our study, the Cronbach was .96.
Nurse-Physician Communication.
Nurse-physician communication was measured by using part of the ICU Nurse-Physician Questionnaire.27 The overall questionnaire consists of 47 scales used to measure variables that affect relationships between nurses and physicians. For our study, only the scales for measuring between-group communication were used. Other investigators have similarly adapted the questionnaire.12 The 4 scales used to measure communication between nurses and physicians were openness (4 items), accuracy (5 items), timeliness (4 items), and understanding of the communication that occurs between nurses and physicians (8 items). Although developed for use in ICU environments, the tool has been recommended for use in other settings and was used recently in a study of medical-surgical nurses.12 Convergent and discriminant validity were assessed through factor analysis.27 Shortell et al27 reported Cronbach reliabilities ranging from .64 to .88. Other researchers have reported reliabilities of .8512 and .86.28 In our study, the Cronbach was .92.
Patient Safety Outcomes. Nurses were asked to indicate on a 4-point Likert-type scale (ranging from never to frequently) how often medication errors, VAP, and catheter-associated sepsis had occurred among patients under the nurses’ care during the previous 3 months. Because research on medication errors "is hampered by a lack of reliable data on the number and type of errors,"21(p214) accurate measurement of medication errors is difficult.29 Although investigators have used chart reviews and incidence reports to gather data on medication errors, direct observation provides the most complete data.22
A plausible alternative to gathering actual data may be the self-report method of determining the occurrence of adverse events, which has been used previously.2,30 In a study30 of nurses’ reports of the quality of care the nurses provided and patients’ outcomes, between-group variance (between groups of nurses on multiple units) was significant, and relationships were correlated in predicted directions (ie, nurses who reported that they provided lower quality of care also reported higher rates of adverse events such as medication errors and nosocomial infections), suggesting that nurses may accurately report information about patients under the nurses’ care.
Data Analysis
Data were analyzed by using an SPSS statistical software program (SPSS Inc, Chicago, Illinois). Descriptive statistics provided information on nurses’ age, sex, ethnicity, education, tenure, status, and years of experience. Information also was collected on the size and type (teaching or nonteaching) of the hospitals, because these factors also influence patients’ safety but often are ignored in nursing research.29 Data were coded before entry into an SPSS database. For checks of coding accuracy, 5% of individual entries were selected randomly and the coded data were compared with raw data. All detected errors were corrected. In order to minimize the likelihood of error, data cleaning was done at the end of each data entry session by reviewing frequency distributions to detect nonvalid responses. The level of statistical significance chosen was .05.
Data were analyzed in 2 ways: first, at the individual level (nurse), because nurses’ perceptions were the focus of interest, and second, at the group level (unit), to determine if any differences were significant when nurses’ responses were clustered by ICU. Random intercept multilevel models were generated to account for clustering and correlated observations at the ICU level. With clustered data, estimated standard errors in a conventional regression are smaller than actual standard errors because standard regression models do not take into account the correlated structure of observations.31 Underestimating standard errors increases the risk of committing a type I error, meaning that the possibility is higher that results will be interpreted as significant when they are not.31 Multilevel models are able to correct the dependency of observations within clusters and thus are appropriate for use with nested data.31
|
Results |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
| As nurses’ perceptions of magnet characteristics and communication with physicians increased, medication errors decreased.
|
A correlation matrix with the Pearson correlation coefficient was used as a screening tool to test for associations between main study variables (Table 1
). Both the workplace empowerment and the practice environment scales were positively and significantly (P < .001) related to the communication scale, suggesting a relationship between work environment factors and communication with physicians. The workplace empowerment and practice environment scales were both strongly associated with each other (r = 0.61, P < .001), probably because of conceptual overlap between the 2 scales. Medication errors had a significant inverse relationship with the practice environment scale and the communication scale. These results suggest that as nurses’ perceptions of magnet characteristics in the practice environment and communication with physicians increased, medication errors decreased.
|
Possible control variables included in the models were nurses’ characteristics, hospitals’ characteristics, and physicians’ communication characteristics. Nurses’ characteristics were age, sex, ethnicity, educational level, years of experience, years in current institution, years in an ICU, mean number of hours worked per week, type of position, and certification in a specialty. The sole hospital characteristic was ICU type. Physicians’ communication characteristics included the amount of time spent communicating with various levels of physicians (first-year resident, second-through fourth-year residents, fellows, attending physicians, patients’ own physicians, and consultants) and the percentage of time that physicians returned a page within 5 minutes. These variables were included in the models because all could possibly be associated with nurse-physician communication.
A series of 2-level random intercept models (level 1=nurse; level 2=ICU) was generated to address each aim. To address the relationship between factors in the ICU work environment and the effectiveness of nurse-physician communication, we considered inclusion of several control variables. Age was the sole significant nurses’ characteristic associated with nurse-physician communication; ICU type and percentage of time that physicians returned a page within 5 minutes also were significant control variables. Table 2 shows the final model. Significant coefficients for workplace empowerment and characteristics of magnet hospitals indicate that both of these work environment factors influenced nurse-physician communication at the level of nurses.
|
To address the relationship between factors in the work environment, the effectiveness of nurse-physician communication, and nurse-assessed outcomes (VAP, catheter-associated sepsis, and medication errors), we entered the same variables of nurses’ characteristics, hospitals’ characteristics, and physicians’ communication characteristics into the models (Table 3). Only years of experience was significantly associated with medication errors, and the relationship was positive, suggesting that the more years of experience a nurse had, the more likely he or she was to report medication errors.
|
|
Discussion |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
| The more years of experience nurses had, the more likely medication errors were reported.
|
The discovery of a link between factors in the work environment and nurse-physician communication also provides some insight into ways to improve nurses’ perceptions of the communication they share with physicians. Providing nurses with more information, support, resources, and opportunities at work can improve communication with physicians. Characteristics of magnet hospitals that can help improve nurse-physician communication include nurses’ involvement in hospital affairs, using a professional nursing model to deliver patient care on a unit, and adequate staffing.
Mohr and Batalden32 assert that healthcare teams need integrated information and support from the larger organization to be effective. Information and support are 2 components of workplace empowerment, and supportive leadership is a characteristic of magnet hospitals, suggesting that the characteristics of the work environment we identified can contribute in at least 2 specific areas where improvements in the work environment might affect nurse-physician communication. The relationship between organizational characteristics and nurse-physician communication is poorly understood, however, and other organizational characteristics probably also affect nurse-physician communication. Rosenstein and O’Daniel33 implied that organizational factors contribute to poor communication when they recommended organizational self-assessment as a strategy to decrease disruptive behaviors by physicians and nurses.
| Use of a professional nursing model to deliver patient care and adequate staffing may improve nurse-physician communication.
|
An intriguing finding in our study is that although workplace empowerment and characteristics of magnet hospitals together were strong predictors of nurse-physician communication, neither was associated with any perceived outcome. This result appears to contradict evidence that indicates just the opposite.2 Although Aiken et al2 found significant relationships between organizational characteristics and nurse-assessed (ie, nurses’ perceptions of) quality of care, they did not query nurses on the nurses’ assessments of specific outcomes. Mitchell et al1 also were unable to show a relationship between ICU structural features and outcomes of severity-adjusted mortality or length of stay. Perhaps characteristics of the work environment are not generic or interchangeable, and some facets of the environment may be more strongly associated with certain outcomes than are others.
| Nurse-physician communication was a small but significant predictor of perceived medication errors only.
|
Nurse-physician communication was a small but significant predictor of perceived medication errors only. This finding is important as well as interesting, because few investigators have examined the relationship between a nursing process and outcome. The fact that neither VAP nor catheter-associated sepsis was significantly related to nurse-physician communication but medication errors were suggests that patients’ outcomes, even those deemed nurse-sensitive, should be investigated individually. Other researchers18,34 also found differences across outcomes.
Our study had several limitations. Although self-rating scales have been used in previous research, they do introduce a potential for recall bias and for underestimation or overestimation of outcomes. Our findings must therefore be interpreted with caution. We used a cross-sectional design and did not track the process of nurse-physician communication over time. Longitudinal study is needed to determine if the tested relationships would hold over time.
The nonexperimental cross-sectional design also limited our ability to establish cause-and-effect relationships. Social desirability response effect bias is the tendency of some individuals to overreport socially desirable personal characteristics and to underreport socially undesirable characteristics. This limitation was minimized by assuring potential respondents of their anonymity and of the confidentiality of individual reports.
Finally, the ordering of instruments in each packet was varied to decrease the likelihood of consistency artifact, which often occurs in research in which self-reports are used.35
|
Conclusion |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
To purchase electronic or print reprints, contact The InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050 (ext 532); fax, (949) 362-2049; e-mail, reprints{at}aacn.org.
Now that you’ve read the article, create or contribute to an online discussion about this topic using eLetters. Just visit www.ajcconline.org and click "Respond to This Article" in either the full-text or PDF view of the article.
To learn more about AACN’s healthy work environment standards, visit www.ajcconline.org and read the article by McCauley and Irwin, "Changing the Work Environment in Intensive Care Units to Achieve Patient-Focused Care: The Time Has Come" (American Journal of Critical Care, November 2006).
This study was funded by the Blue Cross Blue Shield of Michigan Foundation. Barry DeCicco was paid as a statistical consultant for the research.
|
REFERENCES |
|---|
|
TopAbstractObjectivesMethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Schmalenberg and M. Kramer Clinical Units With the Healthiest Work Environments Crit. Care Nurse, June 1, 2008; 28(3): 65 - 77. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
