While emphasis has been primarily on the effect of educational preparation on technical performance, other variables, such as age, experience, and knowledge, need to be investigated. The major issue in this study was the examination of variables which may explain variance in the technical competence of RNs in administering oral medications. The inclusion of six variables was an improvement over other studies that investigated levels of education as a predictor of performance.
Review of Literature
Few studies have been conducted to examine the difference in clinical competence of registered nurses. The research that has been undertaken has been controversial and contradictory in its findings. A study by Reichow and Scott (1976) showed significant differences among associate degree nurses (ADNs), baccalaureate degree nurses (BSNs), and diploma nurses in leadership and technical skills. They concluded:
1. The diploma nurse and the BSN were equal in leadership skills;
2. The diploma nurse was stronger in technical skills than the BSN; and
3. The ADN was not strong in any area.
Nelson (1978) also identified differences in clinical competence among the three types of graduates. However, her findings were inconsistent with the conclusions of Reichow and Scott. Baccalaureate degree nurses rated themselves higher in communication skills than ADNs and diploma nurses, while diploma nurses felt they were superior to BSNs and ADNs in technical and administrative skills. By contrast, nursing supervisors of the respondents rated the baccalaureate nurses higher than other nurses in communicative, technical, and administrative skills. Nelson supported the findings of earlier researchers (Davis, 1966; Dyer, Monson, & Drimmelen, 1975) who concluded that the clinical performance of the BSN was superior to that of the ADN and diploma nurse.
Waters, Vivier, Wilson, Chater, and Urrea (1972) utilized a slightly different approach for comparing ADNs and BSNs. The researchers interviewed staff nurse subjects and administrators instead of relying on written self-reports or supervisor ratings. The researchers concluded that the nursing actions of ADNs were consistent with technical nursing practice. The head nurses of the staff nurse subjects and 10 of the 12 directors of nursing interviewed reported specific differences between the two nurse groups. Associate degree nurses were described as being mechanical and "hospital-oriented." Baccalaureate degree nurses, on the other hand, could provide comprehensive care with expertise in communication, teaching, and use of community resources.
While the previously discussed studies focused on the skills of nurses, several studies were conducted to examine differences in state board scores among nurses. In 1969, Bruegel reviewed the State Board results of four nursing programs in Colorado and found no significant differences in results among educational levels. Zarett (1980), on the other hand, reviewed the 1979 State Board results of all nursing graduates in Pennsylvania, and found that the diploma graduates scored significantly higher than the BSN graduates in 6 out of 11 nursing areas. The AD graduates rated the lowest in all areas.
While most of the cited researchers used written questionnaires and rating scales as their tools for measuring nursing competence, this study utilized direct observation of nurses from diploma, AD, and BS nursing programs. In addition to educational level as a basis for comparison, this research included other variables such as age, experience, and knowledge of the nurses in the sample.
This study addressed the following questions:
1. Does the type of basic nursing education program explain the variance in the technical competence of nurses in the administration of oral medications?
2. Does drug knowledge explain the variance in the technical competence of nurses in the administration of oral medications?
3. Do selected demographic factors (age, RN experience, previous nursing experience, and highest nursing education completed) explain the variance in the technical competence of nurses in the administration of oral medications?
4. Do basic education, drug knowledge, and selected demographic variables explain the variance in the technical competence of nurses in the administration of oral medications?
Subjects: Sixty female nurses, 16 ADNs, 17 BSNs, and 27 diploma nurses, volunteered to participate in the study. Seven of the nurse subjects were more than 40 years of age. All of the subjects were employed in medical-surgical areas of a community hospital on the East Coast.
Research Variables: The criterion variable was the technical competence of RNs in the administration of oral medications. The predictor variables were basic nursing education, drug knowledge, age, RN experience, previous nursing experience, and highest education completed.
Technical competence was measured by observing subjects to obtain total competence scores on the Technical Competence Tool, indicating skill in the hospital-standardized procedure for oral medication administration.
Basic nursing education was measured by asking each subject to indicate whether her basic preparation to become an RN was obtained in an associate of arts degree program, a diploma program, a baccalaureate degree program, or "other" type of program.
Drug knowledge was measured by asking each subject to complete a Drug Knowledge Profile to obtain a score which indicated the knowledge of each RN.
Age was measured by asking each subject to indicate whether her age was between 19-24 years, 25-29 years, 30-39 years, 40-49 years, 50-59 years, or over 59 years.
RN experience was measured by asking each subject to indicate whether she had less than one year of experience as an RN, 1-4 years, 5-9 years, 10-14 years, 15-20 years, or more than 20 years.
Previous nursing experience before becoming an RN was measured by asking each subject to indicate whether she had previous experience as a nursing assistant, LPN or LVN, no experience, or "other" type of experience.
Highest education completed for a nursing degree or diploma was measured by asking each subject to indicate whether she had obtained an associate of arts degree, a diploma, a baccalaureate degree, or "other" degree or diploma.
Instrumentation: Three instruments were developed for the collection of data: a tool for measuring technical competence (TCT), a Drug Knowledge Profile (DKP), and a Demographic Data Sheet (DDS).
Technical Competence Tool: The Technical Competence Tool (TCT) was a checklist containing 16 critical behaviors necessaray for accurate and safe oral medication administration. Each nurse subject received a "yes" or "no" for each of the 16 items to indicate the accomplishment of the critical behaviors for administering oral medications. A value of 1 was assigned to each "yes" response and a 0 was given for each "no" response. A score was then tabulated for each of the two medications administrated by each subject, and the two scores were summed to obtain a total score. Since each nurse was observed at two different times, the total score for observation #1 and that for observation #2 were summed to obtain a total competence score. This resulting score ranged from 0-64, with high scores indicating a high level of technical competence in administering oral medications.
Content validity of the TCT was assessed by three experts in nursing. One taught pharmacology to nursing students in a diploma program and two taught in a hospital inservice education department. All three persons were responsible for constructing pharmacology tests for nursing students or nursing graduates. The experts were asked to review each item of the TCT to determine if it measured the objective that it was intended to measure. The primary researcher then computed the Index of Item-Objective Congruence as a measure of item content validity (Martuza, 1977).
Since there were two observers measuring the performance of subjects, inter-rater reliability was estimated. The assistant observer (rater #2) was a BSN and part-time graduate student in nursing. Nine of the 60 subjects were observed by the assistant. During the piloting of the TCT, the observers trained until a .85 reliability coefficient was obtained.
Intra-rater reliability was assessed by computing a Pearson Product-Moment Correlation between medication #1 and medication #2 scores for each observation. Correlations between the total scores from observation #1 and observation #2 were also computed.
The Drug Knowledge Profile: The Drug Knowledge Profile (DKP) was a 14-item cognitive test designed to measure drug knowledge. All of the items were multiple choice questions. A numerical value of 1 was given for each correct response and a value of 0 was given for each incorrect response.
Two nursing instructors, who were responsible for pharmacology test construction in a school of nursing or inservice education department, were asked to assess content validity of each item, as done with the TCT. The ItemObjective Congruence Index was utilized to quantify content validity of the DKP. The alpha coefficient (Martuza, 1977) was used as an estimate of internal consistency.
The Demographic Data Sheet: The Demographic Data Sheet (DDS) was a five-item tool used to collect information regarding age, education, and experience of the subjects. Responses on the DDS were coded and scored as frequencies.
Procedure: A pilot study was conducted to pretest all three tools in a hospital other than the research agency. This institution was twice the size of the research setting, but had a similar cross-section of registered nurses, i.e., approximately equal distribution of ADNs, BSNs, and diploma nurses. The ages and work experience of the nurses in the pilot setting were approximately the same as those of the nurses in the research institution.
A small group of nurses was asked to participate in the pilot study to aid the researchers in the clarification of the instruments and in the estimation of the time required for each subject to participate. Of the eight nurses asked to volunteer, no one refused.
After the results of the pilot were reviewed and the appropriate revisions of the instruments were made, the data collection for the major study began. Each nurse subject was observed as she administered two medications to one patient. The Demographic Data Sheet and the Drug Knowledge Profile were completed by the nurse after the first observational period. Anonymity of the nurse subjects was preserved at all times.
Within three weeks after the first observation, each subject was again observed as she administered two oral medications to a different patient. It was thought that two observational periods would help to reduce the Hawthorne effect which might be present during the procedure. Morgan and Irby (1978) stated that multiple observations, in a natural setting, should be conducted to check for consistency of performance. The study was conducted over a five-month period.
Limitations: There were several limitations to the research. First, the Hawthorne effect might have prevented the attainment of accurate competence scores. Even though the subjects were asked to give the medications as they usually did, they could have tried to perform the procedure more carefully when observed.
The second limitation was concerned with the size, location, and the nature of the sample. The variable to subject ratio was 1:10, which is lower than the desired 1:30 for a regression study. The significance of the findings might have been limited by the use of a convenience sample rather than a random selection or assignment.
Although the sample was 60 subjects, all of these nurses were employed in the same hospital, forming a homogeneous population. In addition, due to time and financial constraints, only one technical skill was examined as a representation of overall technical competence.
Results and Discussion
Reliability and Validity of the Instruments: The Index of Item-Objective Congruence to quantify content validity of the Drug Knowledge Profile (DKP) was 1.00 for each of the 14 items, demonstrating perfect item-objective congruence (Martuza, 1977).
Using alpha, the internal consistency of the DKP was estimated as .53 for the pilot and .43 for the major study. According to Martuza (1977), if the items of a test are not homogeneous, a relatively low internal consistency coefficient will result. Although all of the items on the DKP measured drug knowledge, there were several groups of questions within the tool, i.e., items which addressed drug interactions, drug classificatms, drug side effects, and drug dosage calculations. When the tool was divided into these categories, however, there was no significant increase in the alpha coefficient for each category. Another factor which could have accounted for the low internal consistency coefficients was the small number of items on the DKP. As a rule, "the longer the test, the more reliable it tends to be" (Martuza, 1977, p. 131). The third, and perhaps most important factor that could have accounted for this result, was the homogeneity of the sample causing a small variance in the results (Var (x) = 4.98 for pilot, and .92 for major study) (Martuza, 1977).
The Index of Item-Objective Congruence to quantify content validity of the Technical Competence Tool (TCT) was 1.00 for each of the 16 items, demonstrating perfect congruence of items with the objective of the tool.
The raters were able to reach a .85 inter-rater reliability after the second subject. The reliability for subjects #1 and #2 were .82 and .91, respectively, during the pilot study. Inter- rater reliability was assessed as .97 during the major study. It was important to establish a high reliability in view of two independent raters.
Intra-rater reliability was estimated as .85 and .77 for rater #l's observations of subjects, and .83 and .78 for rater #2's observational periods. In both groups (rater #l's and rater #2's), the intra-rater reliability coefficient was smaller for observation #2. This was not expected as it was thought that rehearsal during observation #1 would cause an increase in the intra-rater reliability during observation #2.
Research Question One: Summary statistics were calculated and a simple regression was performed to determine if competence scores were a function of basic nursing education. The total competence scores ranged from a low of 37 to a high of 58 (of 64 possible points). The overall mean was 46.33. Five of the 16 critical behaviors for oral medication administration were often omitted by the nurse subjects: items #1, 7, 8, 13, and 15.
Item #1, which states "Washed hands prior to administering medications," and item #7, which states "Rechecked medication labels with medication book after pouring medications," were deleted most often by the subjects. Several nurses told the rater that they realized that they should wash their hands, but were too busy to do so. Since the medications were unit-dose packaged, many nurses did not feel it necessary to recheck the medication labels.
Item #8 (checking arm bracelets) was performed by less than one third of the subjects. Several nurses told the rater that they recognized the importance of checking arm bracelets, but felt that they were familiar with the patients to whom they were giving the medications. About one-half of the subjects did not place their initials in the appropriate space on the medication sheet after the medications were administered (item #13). Instead, they signed their initials before giving the medications. Approximately two thirds of the subjects wrote their signature on the medication sheet (item #15).
As seen in Table 1, the baccalaureate educated group achieved the single highest (58) and lowest (37) competence scores on the TCT. The AA degree group obtained the highest mean score on the basic education variable.
The results of the regression analysis showed that less than one pereent of the variance in the competence scores could be explained on the basis of nursing education.
Research Question Two: Summary statistics and a simple regression were utilized to determine if competence scores were a function of drug knowledge. The scores on the DKP ranged from 10 to a perfect score of 14. The overall mean score was 13.06. The regression analysis showed that less than one percent of the variance could be explained by drug knowledge.
Research Question Three: Summary statistics and a multiple regression analysis were calculated to determine if competence scores were functions of selected demographic factors. The nurses in the 25-29 age group scored higher on the TCT than any other age group. Subjects with one to four years of RN experience had the highest TCT mean score in the sample. As expected, those RNs who had previously been LPNs scored higher on the TCT than those who had only been nursing assistants. However, the nursing assistant group scored higher than those subjects who had no previous exposure to nursing before becoming an RN. The examination of the highest nursing education completed variable revealed no significant difference from the basic education results, i.e., the AA degree group obtained the highest mean scores on the TCT
SUMMARY STATISTICS FOR TOTAL COMPETENCE SCORES PER GROUP BY BASIC NURSING EDUCATION
Two percent of the variance in total competence scores was explained by age, but the remaining demographic factors accounted for little more than one percent. The largest beta weight was for age, but the standard error in relation to the size of the beta weight was also large.
Research Question Four: Multiple regression analysis was utilized to answer the fourth research question. All six predictor variables were examined to determine their contribution to total competence scores as shown in Table 2. Approximately three percent of the variance was explained by the six predictor variables when analyzed as a group. All significance was tested at ? =£ .05.
Although selected predictor variables explained less than three percent of the variance in competence scores, the findings from this study suggested that the technical competence of RNs administering oral medication varies.
These findings were similar to those of Reichow and Scott (1977), and Nelson (1978) who found that RNs varied in their performance of technical skills. Unlike these studies, however, this research did not support basic nursing education as a predictor of performance.
In 1972, Kohnke found that nursing school curricula were often not closely aligned to the literature description of those curricula. This study would seem to support her findings as there was no significant difference in the performance of AD, diploma, or BS graduates in the administration of oral medications.
Davis (1966) concluded that work experience was related to performance, as she found that the quality and quantity of nursing care decreased as nursing experience increased. In Dyer et al. (1975), performance decreased when age increased, and increased as the educational level increased. Contrary to the findings of these researchers, the results of this study did not support nursing experience or age as predictors of nursing competence.
RESULTS OF FORWARD STEPWISE REGRESSION ANALYSIS OF PREDICTOR VARIABLES ON TOTAL COMPETENCE SCORES (N = 60)
Significance for Nursing Research
Research, such as this study, is needed to support or refute the delineation of the "technical" and "professional" nurse, as specified by Montag in the 1950s and the American Nurses' Association for 1985. These findings are vital as information for the consumer, for other health professionals, and for the nursing profession itself.
This research could also provide meaningful information to schools of nursing, as they continue to revise their curricula to meet the demands of the employer and health care consumer. With a nationwide increase in nursing-related medication errors, a need exists to identify the factors which might be contributing to these incidents (Trends, 1980). The procedure being taught and/or used by nurses for medication administration might need review and revision.
On the basis of the findings of this study, it is recommended that:
1. This study be replicated incorporating the following changes designed to overcome the limitations identified:
a. the use of a larger sample,
b. the use of a random sample utilizing more than one hospital, and/or
c. the inclusion of additional predictor variables, such as length of employment, staffing patterns, and patient assignment.
2. A study be undertaken to examine the same technical skill, i.e., administration of oral medications, utilizing a larger convenience sample in a metropolitan teaching hospital.
3. A study be conducted to examine another technical skill, e.g., urinary catheterization, sterile dressing change, or intramuscular injection, utilizing a larger convenience sample in the same research setting.
4. A study be conducted to examine another technical skill, as suggested earlier, utilizing a larger, randomly selected sample from several hospitals.
- Davis, F. (Ed.). (1966). The nursing profession: five sociological essays. New York: John Wiley & Sons.
- Dyer, E., Monson, M.A., & Drimmelen, V. (1975). What are the relationships of quality patient care to nursing performance, biographical and personality variables? Psychological Reports, 36, 255-266.
- Martuza, V. (1977). Applying norm-referenced and criterion-referenced measurement in educational research. Boston: Allyn and Bacon.
- Morgan, M.K., & Irby, D.M. (1978). Evaluating clinical competence in the health professions. St. Louis: CV. Mosby Co..
- Nelson, L.J. (1978). Competence of nursing graduates in technical, communicative and administrative skills. Nursing Research, 27, 121-125.
- Reichow, R. W, & Scott, R.E. (1976). Study compares graduates of two-, three-, and four-year programs. Hospitals, 50, 95-97.
- Trends. (1980). Nursing '80, 10, U.
- Waters, V.H., Vivier, M.L., Wilson, H.G., Chater, S.S., & Urrea, J.H. (1972). Technical and professional nursing: An exploratory study. Nursing Research, 21, 124-131.
- Zarrett, A (1980). Is the BSN better? RN, 78, 28-33.
SUMMARY STATISTICS FOR TOTAL COMPETENCE SCORES PER GROUP BY BASIC NURSING EDUCATION
RESULTS OF FORWARD STEPWISE REGRESSION ANALYSIS OF PREDICTOR VARIABLES ON TOTAL COMPETENCE SCORES (N = 60)