| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Corresponding author: Barbara B. Brewer, RN, PhD, MALS, MBA, 1442 E Marco Polo Rd, Phoenix, AZ 85024 (e-mail: barbara.brewer{at}jcl.com).
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionSummaryReferences |
|---|
• Objectives (1) To test application of a case report form used to assess patients’ characteristics as defined by the American Association of Critical-Care Nurses (AACN) Synergy Model for Patient Care in a general population of pediatric and adult patients and (2) to evaluate the internal consistency reliability and construct validity of the patient characteristics measure found on the American Association of Critical-Care Nurses Web site.
• Methods A cross-sectional correlational study was conducted in 2 phases. The first phase consisted of secondary data analysis of 481 ratings of patients provided by 11 expert nurses. The second phase consisted of primary data collection of 279 ratings of patients provided by 116 general and critical care nurses. The case report form was used to rate characteristics of patients in both phases; a self-rated nursing proficiency scale was used in the second phase. Descriptive statistics were used to describe the sample. Correlational techniques were used to evaluate internal consistency reliability and evidence of construct validity.
• Results The case report form based on the AACN Synergy Model’s characteristics of patients showed satisfactory internal consistency reliability and evidence of discriminant construct validity. Exploratory factor analysis resulted in a 2-factor solution representing an intrapersonal interaction factor and an interpersonal interaction factor.
• Conclusion The case report form for assessing characteristics of patients showed utility in a general population of adult and pediatric patients, some critically ill and some not. Nurses without previous knowledge of the AACN Synergy Model were able to apply the model during routine patient care.
The AACN Synergy Model describes a relationship between a patient and a nurse that acknowledges the importance of nursing care that is based on the needs of patients and their families. Although thorough conceptual analyses were not conducted (M. A. Q. Curley, RN, PhD, oral communication, November 11, 2005), consensus was developed around 8 definitions for specific characteristics of patients and competencies of nurses (Table 1
). Characteristics of patients were believed to reflect the universal needs of patients, whereas competencies of nurses were believed to be attributes necessary among critical care nurses to ensure optimal outcomes for patients.2
|
| The AACN Synergy Model matches patients’ characteristics with nurses’ competencies to optimize outcomes.
|
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionSummaryReferences |
|---|
). The form was modified to include a self-rating component (Figure 2) based on Benner’s model5 and brief demographics.
|
|
Although most units at the study site were traditional critical care or general care units, one unit was an acuity-adaptable, universal unit where patients’ conditions ranged from critical to stable or ready for discharge. The purpose of the second phase of the study was (1) to evaluate the ability of nurses unfamiliar with the AACN Synergy Model to complete the case report form (utility) and (2) to evaluate whether construct validity findings in this sample would support those from the earlier study.
Expert rater data were collected in 2003 during a pilot study intended to measure the relationship between the characteristics of patients as defined in the AACN Synergy Model and acuity scores; because this project was never completed, pilot data had been archived and had not previously been analyzed. The project leaders determined "expert" status at the time of the pilot study. All expert nurses were deemed experts in application of the Synergy Model because they had read extensively about the model and had studied the previous 2 years with Dr Martha A. Q. Curley. In addition, they had incorporated the model’s competencies of nurses into the healthcare system’s program for advancement of nursing careers. Data from the pilot study on patients’ characteristics were collected during a 5-day period by the expert raters after observation of the patients and interviews with the patients’ nurses. Interviews focused on specific physical, psychosocial, and care process findings aligned with the Synergy Model’s characteristics of patients. We conducted secondary analyses on data from this arm of the study.
The distribution of patient care units during the second (naive rater) phase of the study was similar to that used in the expert rater phase, but data were collected prospectively on day shifts for 5 days in the spring of 2005. Sample size was set at 30 patients per unit. Naive rater nurses were asked to complete a brief demographic survey, to self-rate their own level of practice according to Benner’s novice-to-expert definitions,5 and to rate each of their assigned patients on the case report form (Figure 1
). Nurses in this phase of the study were considered naive because they had not been schooled in the AACN Synergy Model.
To ensure that time was sufficient for assessment and interaction with patients before the rating was obtained, all nurses completed the case report form after the midpoint of their shift. Completed case report forms were placed in sealed envelopes by each participating nurse, then collected at the end of each shift by the coinvestigators. Only the coprincipal investigators had access to the raw data, and they took responsibility for all data input and analyses.
In both the expert and naive rater phases, nurse participation was voluntary and no data identifying the patient or the nurse were collected. Data were analyzed with SPSS version 12.0 (SPSS Inc, Chicago, Ill); descriptive statistics and bivariate correlation coefficients were computed for all variables. Underlying factor structure was estimated by using principal components extraction with varimax rotation. Statistical significance was set at P < .05.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionSummaryReferences |
|---|
presents the 8-item, 5-level case report form used to assess patients’ characteristics. The instrument, which was adapted from an existing measure found on the AACN Web site,4 had satisfactory internal consistency reliability, with a Cronbach 
of .88 in both the expert and naive rater phases of the study. Lower scores on each of the 8 characteristics indicate greater criticality. For example, a score of 1 on the vulnerability characteristic indicates a patient who is highly vulnerable.
Rater Characteristics and Sample
Expert Rater Phase.
A total of 11 nurse raters collected data in the first phase of the study. Of these, 10 had at least 10 years of practice experience, and 1 had 5 years of experience. All 11 nurses were considered experts in clinical practice and on the operational definitions for characteristics of patients from the Synergy Model (Table 1). Expert raters completed a total of 481 case report forms on patients from less than 1 year to 95 years old (mean 51.8 years; SD 23.2). Table 2 shows the distribution of the sample by unit.
|
). Naive raters had from less than 1 year to 37 years of experience (mean 9.8 years, SD 9.5, median 6) and had worked a mean of 4.8 years (SD 5.6, median 2.5) in their currently assigned unit. Naive raters classified their level of clinical nursing expertise (Figure 2) as "proficient" on average; 59% had attained a baccalaureate degree, 32% an associate’s degree, and 5% a registered nursing diploma; 4% of the case report forms contained no data on level of education.
| Nurses untrained in the AACN Synergy Model were able to rate the 8 patient characteristics; no redundancy among these items was found.
|
Item Redundancy Testing.
Interitem correlations, means, and SDs for ratings of patients’ characteristics from each study phase are presented in Table 3. Item redundancy was not found; interitem correlations ranged from 0.22 to 0.75 in the expert phase and from 0.29 to 0.78 in the naive phase. In both phases, resource availability had the weakest correlations with the other characteristics of patients. Mean scores obtained by expert raters were lower, suggesting worse health status (eg, more vulnerable, more complex, less stable conditions) among this cohort than among the naive raters’ sample of patients.
|
Examination of the data base revealed that 4 raters generated all of the missing scores. Two highly experienced nurses (25–30 years of experience) who worked in an orthopedic unit did not provide the 3 missing resource availability ratings and 1 of the missing predictability ratings. An inexperienced nurse (<6 months of experience) who worked in a pediatric hematology oncology unit did not provide a predictability rating. An experienced nurse (30 years of experience) who worked in an adult blended acuity unit did not provide a stability score.
Construct Validity
Evidence for construct validity was evaluated by using 2 methods. Exploratory factor analysis with principal components extraction with varimax rotation was used to evaluate the underlying factor structure/dimensionality of the 8 characteristics of patients in both phases of the study. Differences among known groups were then evaluated to find evidence for discriminant construct validity.
Factor Analyses.
In both the expert rater and naive rater phases, analyses resulted in a 2-factor solution explaining 71% of the variance in the expert phase and 69% of the variance in the naive phase (Table 4). Five items (vulnerability, stability, complexity, resiliency, predictability) loaded on the first factor during both study phases; this factor was named the intrapersonal interaction factor because it loaded items that the investigators associated with characteristics primarily driven by the patient’s internal being. Three items (participation in decision making, participation in care, resource availability) loaded on the second factor; this factor was named the interpersonal interaction factor because it loaded those characteristics that the investigators associated with the patient’s interface with the external environment.
|
In the expert rater sample, types of patients differed significantly for each of the 8 characteristics of patients (F > 10.6, P < .001). Table 5 provides mean scores by type of patient and results of post hoc comparisons for group differences. In the expert rater sample, mean scores for each of the 8 characteristics were significantly lower in critical care patients than in patients in both acuity adaptable/universal units and general care units. In the naive rater sample, only 3 characteristics (vulnerability, resiliency, and participation in care) were rated significantly differently (F > 3.93, P < .05) between critical care patients and the 2 other acuity levels. Mean scores for these 3 characteristics were significantly lower in critical care patients than in patients in both acuity adaptable/universal units and general care units. For the most part, ratings for patients in acuity adaptable units fell between the scores for patients in critical care and general care units. In all cases, scores were lower for critical care patients than for patients in general care units.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionSummaryReferences |
|---|
We did not ask raters to provide feedback about ease of use of the measure. As a result, we do not know whether missing data were due to confusion about the item, inability to provide a rating because of lack of knowledge about the particular characteristic, or some other reason. We also cannot comment on consistency among raters or accuracy of the ratings because we did not assess interrater reliability or collect data that would allow us to evaluate concurrent validity. Further research is needed to answer these questions.
Ratings of patients’ characteristics accounted for 71% and 69% of state variance when used by expert and naive raters, respectively. This finding is important and provides a foundation for further research into use of the model to complement measurement of patients’ acuity, documentation methods, and, ultimately, alignment of nursing expertise with key findings related to patients’ characteristics.
Most likely, some findings in our study reflect the diversity of the 2 samples of patients as well as the variation in practice experience, level of education, and self-rated level of expertise among our naive raters. Use of 2 different groups of patients (ie, one group rated 2 years earlier by experts and another group rated later by naive nurses) limits our ability to compare measures provided by expert and naive raters directly with each other, and the patients in the 2 samples may have differed somewhat in their health status. For example, we are unable to determine if the finding among naive raters that only 3 characteristics of patients differed significantly in their scores by type of unit (eg, critical care vs noncritical care units) was due to actual differences in patients’ characteristics or the expertise of the naive nurse sample.
Additionally, our naive raters predominantly rated themselves as proficient, indicating a more concrete approach and ease in identification of readily repeatable patterns among patients. Had patients in the naive rater sample varied in the findings associated with characteristics other than vulnerability, resiliency, and participation in care, ready identification and rating of these findings could have been missed, thereby accounting for the limited differences found. Interestingly, this result would be consistent with results seen in Benner’s work6 and validates the need for expert clinical resource nurses (eg, advanced practice nurses) to support the early identification and treatment of health problems.
To date, comprehensive concept analyses of the Synergy Model’s characteristics of patients have not been published. However, the original definitions proposed for each characteristic were adequate for completion of patients’ ratings by the naive users in our study. We do think that thorough conceptual analyses would further enhance the model’s utility, because even among ourselves we have found discrepancies in definitions that were best resolved through sharing of exemplar case studies, critique of related literature, and dialogue. In addition, should the model be tested in settings other than traditional acute care hospitals (eg, psychiatric mental health, outpatient clinics), concept analyses would strengthen application, because most likely these characteristics of patients can be applied in settings other than acute care hospitals.
Similar to Benner’s work,6 the AACN Synergy Model proposes that to support optimal health outcomes, nurses’ competencies should match the characteristics of patients in healthcare settings. The competencies of nurses defined in the AACN Synergy Model have yet to undergo similar scrutiny, and the design of methods to measure nurses’ competencies will have to be carefully orchestrated to ensure the confidentiality of the nurse subjects. If such an undertaking is successful, and if it can be matched with methods to quantify patients’ characteristics, optimal assignment matrices may be developed to support improved systems for delivering patient care that are designed with both the patient and the caregiver in mind. Such an approach would be consistent with the recommendations put forth by the Institute of Medicine for health professionals’ practice and education7 and would increase the value and visibility of nursing.
| These data provide one step toward measuring the effect of nursing practice on patients’ outcomes.
|
|
Summary |
|---|
|
TopAbstractMethodsResultsDiscussionSummaryReferences |
|---|
In both studies reported here, the case report form used to assess the characteristics of patients was reliable. Support for evidence of construct validity in the expert nurse group was shown. Support for evidence of construct validity in the naive nurse group showed mixed results. Future research should include measurement of interrater reliability by having expert and naive nurses evaluate the same patient. Additionally, future research should include correlation of patients’ ratings to their clinical condition, which would provide support for criterion-related validity. Both of these factors are important in helping us understand whether differences in ratings reflect measurement error or actual differences among patients.
|
ACKNOWLEDGMENTS |
|---|
FINANCIAL DISCLOSURES
None reported.
* See Acknowledgments section for current author affiliations.
To purchase 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.
|
REFERENCES |
|---|
|
TopAbstractMethodsResultsDiscussionSummaryReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
