Nursing diagnosis : educational strategy based on problem-based learning 1

This study aimed to evaluate the effectiveness of a teaching strategy on nursing diagnoses based on problem-based learning in the performance of clinical reasoning and diagnostic trial of undergraduate students. Experimental study conducted in two phases: content validation of problems and implementation of educational strategy. The results showed improvement in the data pooling ability of students in the experimental group. It is concluded that the strategy put in practice exerted a positive influence.


Introduction
The decision making process in health requires that nurses be capable of thinking critically and choosing adequate actions to solve the identified problems (1) .In this process, learning clinical reasoning is one of the most complex phases in cognitive terms and one of the themes most directly involved in clinical decision making.
In this context, teaching on the nursing diagnosis process refers to the processing of clinical assessment data, to the establishment of relations among them and to the identification of care needs.Moreover, in this process, students' active participation is stimulated, so as to permit further knowledge on nursing practice (2)(3) .
On the other hand, some authors defend that nursing diagnosis teaching has been conducted without providing students with critical reflection, in view of the need to develop strategies focused on gaining diagnostic skills (4)(5) .The following facts different authors mention support these ideas: diagnostic teaching is not stimulated and addressed as early as the first course subjects: there is a lack of knowledge on the taxonomies used and on the physiopathology of diagnoses; students' difficulty to diagnose based on a nursing reference framework, differently from the biomedical model; incipient diagnosis use in practice during academic life; and some teams, nurses and faculty members' lack of interest, which may originate in knowledge deficiencies and non understanding of the dimension diagnostic use can represent for the transformation of nursing practice (2,3,(6)(7)(8) .It is highlighted that, although few studies exist that involve PBL and nursing diagnoses, research exists in other knowledge areas, such as medicine (9) , physical therapy (10) , law (11) , geography (12) , among others.

Although proposals have been described in
PBL is a method of learning through discovery which values learning how to learn.It starts with clinical situations that require a decision and tries to guide students to find solutions by learning the skills and competencies needed to identify the best actions for the problems presented (13) .
Thus, this research aimed to assess the efficacy of a teaching strategy on nursing diagnoses based on problem based learning in undergraduate students' performance of clinical reasoning and diagnostic judgment.

Methods
This experimental study was organized in two phases.The first involved problem validation and the second the application of the educative strategy on nursing diagnoses, based on problem based learning.
In the first phase, the goal was to develop and validate the problems that guided the application of the educative strategy, which represented the study focus.
The researchers developed four problems in the form of clinical cases.It is highlighted that, in this study, the terms "case studies" and "clinical cases", found in literature, were called "problems", in line with PBL recommendations.
In this problematizing approach, the problem constitutes a written description of the phenomena or events in reality that stimulate group learning, the activation of previous knowledge and identification of possible solutions.The problem is the fundamental component, the starting point and guideline of the learning process (14) .
The problems the authors developed were clinical situations that covered seven NANDA International (15) domains, which were: Nutrition, Elimination and exchange, Activity/rest, Perception/cognition, Safety/ protection, Comfort and Growth/development.
Regarding content validity, the four problems were submitted to a panel of five experts, selected through criteria adapted from Fehring's (16) scoring system.Thus, the population comprised nurses with professional qualifications at M.SC.or Ph.D. level in Nursing.In this study, experts were considered as nurses who scored at least five points on the following criteria (Figure 1).It is highlighted that, in the expert selection, both availability to participate and clinical ability for diagnostic inferences were taken into account.Among the selected experts, one scored nine and four eight.
Content validity of the problems was estimated through the mean inter-expert Content Validity Index -CVI).The CVI was the mean index calculated by adding up the items all experts attributed, divided by the total number of experts.Through this method, the items and instruments on the whole are considered valid if the CVI amounts to 0.90 (17) .CVI's for the four problems Lira ALBC, Lopes MVO.
In the construction of the educative strategy, the goal was to develop a course that provided students with knowledge construction through individual and group analysis of problems.With a view to reaching this goal, PBL was used, in line with the idea (13) that this method, based on significant discovery-based learning, values learning how to learn and intentionally works with problems.To assess the effectiveness of the course, the researchers decided to use a strategy similar to that of the problems discussed during the course modules, with a view to continuing coherence with PBL principles.
Thus, course effectiveness was assessed through a pre and post-test, based on the scores the 30 students obtained on each test.For the pre and post-test, two problems were used that had been validated in an earlier study (18) . The

Results
Study participants were thirty students in the fifth semester of the undergraduate program in Nursing.
None of the students dropped out during the research.Table 3 presents the analysis of differences between the mean pre and post-test scores, between and inside the experimental and control groups.Lira ALBC, Lopes MVO.

Next, in
Scores ranged from zero to ten.When comparing the mean scores, no statistically significant difference was identified between the experimental and control groups (p = 0.406).On the post-test, however, the experimental group's mean scores were statistically higher in comparison with the control group's mean scores (p < 0.001).Intragroup assessment of the mean pre and post-test scores showed a statistically significant difference for experimental group scores only (p=0.001).

Discussion
As for socio-demographic data, the results showed that similar profiles for experimental and control group students before the educative intervention.That minimizes one of the limitations described in one study (19) , in which the sample comprised a disproportionally higher number of students from the final two semesters of the nursing course.According to the same author, the presence of students with different experience or skill levels could affect the study results.Initial uniformity between groups is also important to control for variations in knowledge and previous experience.
On the pre-test, high frequencies were found for the identification of the Excess fluid volume diagnosis among students in both groups.This can be explained by the fact that frequent discussions existed on the defining characteristics involved in this diagnosis -such as edema, blood pressure, breathing, jugular vein turgor, among others -during theoretical and practical classes as part of the Semiology and Physiology subjects.
On the post-test, students from the experimental group evidenced the Risk for infection diagnosis and its risk factors more frequently than in the control group.
In a study accomplished in Japan (20) , two written case studies were used to verify 376 nurses' diagnostic reasoning competency.The study showed that 50% of these nurses correctly identified the Risk for infection diagnosis and 84% its risk factors.According to the authors, that diagnosis is used very frequently in Japan.
In the Brazilian reality, the Risk for infection diagnosis is also discussed quite frequently, mainly when the students learn the nursing procedures that need antisepsis and asepsis care, such as urinary probing for example, during the Semiology subject.
These study results show that the post-test identification of defining characteristics among students from both groups was similar and with low frequencies.
One justification for the small number of defining characteristics students in the two groups identified was that defining characteristics in problem B used for the post-test were written differently from those presented in Nanda International's Taxonomy II; the researchers believe this may have jeopardized the understanding of incipient students in the nursing diagnosis elaboration process.
In a study involving 55 nursing students, aimed at analyzing the relation between cognitive development, clinical knowledge and clinical experience, two clinical cases from the maternal-infant area were used (19) .
The students did not manage to identify many nursing diagnoses.In clinical cases 1 and 2, there were 16 and 11 answer possibilities, respectively.The students identified a mean 3.27 on the first simulation and 1.43 on the second.These data are similar to the present study findings.
In the groups under assessment, a statistically significant difference was perceived between experimental group students' scores on the pre and post-test.On the other hand, in the control group, no such statistically significant difference was found between mean scores on both tests.
Also similar to the present study, in a study accomplished in São Paulo (21) , students from the experimental group showed a significant improvement during the post-intervention phase for the three activities assessed (identify relevant data, group data and name the grouping) in comparison with control group participants.This suggests that the course exerted positive influence on students' diagnostic reasoning.
As the main study limitations, the small sample should be highlighted.Regarding the course, the period it took place and its intensive nature may have entailed different conditions for students in both groups.

Conclusion
This study permitted assessing the efficacy of an This study contributed by using PBL as a teaching strategy to help and establish undergraduate students' clinical reasoning and diagnostic judgment.PBL allows students to construct their own knowledge through the analysis of problems in tutorial groups as well as individually.
literature on nursing diagnosis teaching, few studies have presented a relation between the diagnostic reasoning process as a problemization process and the search for solutions for clinical decision-making.In this respect, Problem Based Learning (PBL) is considered a teaching strategy that can contribute to support the Rev. Latino-Am.Enfermagem 2011 July-Aug.;19(4):936-43.establishment of the relation between diagnosing and decision-making.

Figure
Figure 1 -Expert selection criteria corresponded to 0.90; 0.90; 0.45; 0.55.Hence, the last two problems needed alterations and were again forwarded to the experts.After reformulations, the problems scored CVI's of 0.90 and 0.95.The second study phase was the application of the educative strategy on nursing diagnoses in July 2009.Study participants were students from a Brazilian public university.Thirty students who complied with the eligibility criteria were invited to participate.All thirty accepted and were randomly distributed across two groups, experimental and control.Therefore, the thirty students were consecutively numbered.Next, a sequence of thirty numbers was written on pieces of paper of the same size, placed in a box.The box was shaken and the first fifteen numbers drawn constituted the experimental group, which participated in the educative strategy.The remainder participated in the control group and only received the standard intervention, which included theoretical and practical classes offered as part of the Semiology course in the first semester of 2009.The inclusion criteria were: being a student in the fifth semester of the undergraduate program in Nursing; having passed the Semiology course at the same college in the first semester of 2009, and accepting to participate in all research phases.The exclusion criterion was: people who did not comply with all study phases.

Five
modules were created during the course, each of which involved a different problem, in accordance with PBL recommendations.In the first module, clinical reasoning and diagnostic judgment were discussed.In the other modules, the problems validated in the first study phase were discussed.The course modules started with the presentation of the problem to the students, who did not have previous access to the information needed.The students worked with the problem, organized in tutorial groups.In the tutorial group, the problem was analyzed and the learning objectives were set.Then, the students started the individual study phase at the university library.The student was responsible for identifying relevant bibliographic material.The module finished with a new meeting of the tutorial group, aimed at solving the problem.At that time, the group members systematically presented the results of their individual study and reviewed and systemized the final proposals for the problem.
educative strategy on nursing diagnoses based on PBL, using a pre and post-test.On the pre-test, the two groups identified the characteristics, diagnoses and related factors homogeneously.On the other hand, as expected, students from the experimental group identified more nursing diagnoses and related or risk factors on the post-test.Although students from both groups identified few defining characteristics on the post-test, one may conclude that the educative strategy on nursing diagnoses based on PBL seems to have enhanced undergraduate students' clinical reasoning and diagnostic judgment, as the results showed improved data grouping capacity among experimental group students.
1 -Expert selection criteria expected frequencies were lower than or equal to five.The chance ratio and its confidence intervals were also calculated.For comparisons of means, the t-test was applied, supposing equal or unequal variances, depending on the result of the application of the Levene test.In case of data asymmetry, Mann-Whitney's test was The Kolmogorov-Smirnov test was applied to the numerical variables to check the normality of numerical data, and the Levene test to check the homogeneity of variances.To assess the variables' independence, the Chi-square test was used.Fisher's exact test was used when the

Table 1
, the students' socio-demographic data are presented.

Table 1 -
Distribution of students according to socio- a mean age of 21.7 years.Regarding socio-demographic data, the experimental and control groups showed similar means (p>0.05).The results of the pre-test application showed that the identification of the defining characteristics among Rev. Latino-Am.Enfermagem 2011 July-Aug.;19(4):936-43.www.eerp.usp.br/rlaestudents in the experimental and control groups was

Table 2 -
Analysis of differences in means between the number of defining characteristics, nursing diagnoses and related factors students identified in the experimental and control groups.Natal-RN, 2009 DC-Defining Characteristic; ND-Nursing diagnosis; RF-Related factor; SD-Standard Deviation; K-S-Kolmogorov-Smirnov test; *t Test supposing equal variances; † Mann-Whitney test.No differences were identified in the mean number of defining characteristics, nursing diagnoses and related or risk factors experimental and control group students identified in the pre-test.On the other hand, students from the experimental group identified a higher mean

Table 3 -
Analysis of differences in mean pre-test and post-test scores inter and intra-experimental and control groups.Natal-RN, 2009 K-S-Kolmogorov-Smirnov test; * t Test supposing equal variances; † t Test supposing unequal variances; ‡ Paired t-test Some difficulties were faced to accomplish this study.The first was related to the development of problems that involved clients' typical clinical situations, which the undergraduate students should be familiar with and have previous knowledge about.The second difficulty was that problem development took a lot of time and required great effort to achieve a range of nursing diagnoses.And, finally, the lack of publications on nursing diagnosis teaching strategies through problem-based learning hampered data discussion.Further research is suggested, with a view to the creation of teaching strategies that address other classification systems, based on a common language, linked with the elements of the nursing process, at undergraduate as well as graduate level.This will cooperate for nursing to take hold of its own knowledge base in an organized way, according to its work method.