Journal of Gerontological Nursing

USE OF RESTRAINTS Changes in Nurses' Attitudes

Darlene Schott-Baer, PHD, RN; Stephanie Lusis, MSN, RN, CS; Kay Beauregard, MSA, RN

Abstract

The literature (Werner, 1989; Ramprogus, 1992) suggests that the use of physical restraints on patients has negative consequences , and environments that make use of alternative or nonrestraint interventions to control patient behavior have positive effects on the patient's level of independence (Cober, 1991). However, little information is provided to nurses about the use of physical restraints or alternative methods to control patient behavior, in spite of the fact that nurses, more often than physicians, initiate the application of restraints in elderly patients (Macpherson, 1990). As evidence supporting the benefits of a restraint-free environment accumulates, the necessity of increasing the nurses' awareness of alternative interventions is a primary concern for institutions. Therefore, the purpose of this pretest/ posttest study was to determine whether the attitudes of Registered Nurses (RNs) toward the use of physical restraints became more positive after changes in hospital policy.

For this study, a positive attitude was conceptualized as the nurses' tend to use restraints less often. The new policy emphasizes alternative methods to control patient behavior and is intended to decrease the use of passive restraints on hospitalized elderly patients. Several methods were utilized to communicate the new policy to staff, including inservices; staff meetings; and videotapes entitled "When You Are Considering Restraints" and "Revised Restraint Policy and Procedure." It was hoped that the use of multiple methods of communication, especially the videotapes of alternate methods, would change nurses' attitudes and clinical practice regarding the use of restraints.

Because of the potential for injury (including attendant legal repercussions) resulting from the use of restraints, attention has focused on a reduced use of restraints with eldeny patients in the hospital setting. Confinement through the use of physical restraints can actually increase patient agitation and the potential for falls (Tinetti, 1992; Werner, 1989). For example, in Tinetti and associates' study (1992), a significantly higher incidence of serious fallrelated injuries was experienced by restrained patients over unrestrained patients in skilled nursing facilities. Since physicians rely on nurses to assess patient behavior, nurses are in a position to recommend alternative methods to control patient behavior.

This study is a partial replication of Scherer and associates' (1991) examination of restraint use and the attitudes of nursing home staff toward the use of restraints. However, in that study, only 14% of the sample of 118 subjects were RNs. The educational preparation of the RN subsample was 59% ADN, 29% diploma, and 12% BSN. The reported results were from the pooled data. Respondents felt that staff needed to restrain patients in order to protect them and felt bad if patients became more upset after restraints were applied. Subjects felt that family members did not have the right to refuse the application of restraints but that patients did. This study differs from Scherer's, in that only RNs were sampled, and the setting was a large, midwestern teaching hospital. In addition, a random sampling technique was used to draw a representative sample from three of the nursing divisions within the institution.

Stilwell (1991) found that 86% of her sample (n = 168) of RNs had fewer than 3 hours of instruction related to the use of restraints on adults. Twenty-nine percent of this sample identified medications as the primary alternative to physical restraint, with sitters (19%) and electronic devices (14%) as additional methods. A disturbing 27% of Stilwell's sample were not aware of alternative methods available within their work settings.

Strumpf and associates (1992) also recognized the need of education to reduce the use of physical restraints. A program aimed at understanding the patients' behavior was developed and included discussions about myths, effects of restraints,…

The literature (Werner, 1989; Ramprogus, 1992) suggests that the use of physical restraints on patients has negative consequences , and environments that make use of alternative or nonrestraint interventions to control patient behavior have positive effects on the patient's level of independence (Cober, 1991). However, little information is provided to nurses about the use of physical restraints or alternative methods to control patient behavior, in spite of the fact that nurses, more often than physicians, initiate the application of restraints in elderly patients (Macpherson, 1990). As evidence supporting the benefits of a restraint-free environment accumulates, the necessity of increasing the nurses' awareness of alternative interventions is a primary concern for institutions. Therefore, the purpose of this pretest/ posttest study was to determine whether the attitudes of Registered Nurses (RNs) toward the use of physical restraints became more positive after changes in hospital policy.

For this study, a positive attitude was conceptualized as the nurses' tend to use restraints less often. The new policy emphasizes alternative methods to control patient behavior and is intended to decrease the use of passive restraints on hospitalized elderly patients. Several methods were utilized to communicate the new policy to staff, including inservices; staff meetings; and videotapes entitled "When You Are Considering Restraints" and "Revised Restraint Policy and Procedure." It was hoped that the use of multiple methods of communication, especially the videotapes of alternate methods, would change nurses' attitudes and clinical practice regarding the use of restraints.

Because of the potential for injury (including attendant legal repercussions) resulting from the use of restraints, attention has focused on a reduced use of restraints with eldeny patients in the hospital setting. Confinement through the use of physical restraints can actually increase patient agitation and the potential for falls (Tinetti, 1992; Werner, 1989). For example, in Tinetti and associates' study (1992), a significantly higher incidence of serious fallrelated injuries was experienced by restrained patients over unrestrained patients in skilled nursing facilities. Since physicians rely on nurses to assess patient behavior, nurses are in a position to recommend alternative methods to control patient behavior.

This study is a partial replication of Scherer and associates' (1991) examination of restraint use and the attitudes of nursing home staff toward the use of restraints. However, in that study, only 14% of the sample of 118 subjects were RNs. The educational preparation of the RN subsample was 59% ADN, 29% diploma, and 12% BSN. The reported results were from the pooled data. Respondents felt that staff needed to restrain patients in order to protect them and felt bad if patients became more upset after restraints were applied. Subjects felt that family members did not have the right to refuse the application of restraints but that patients did. This study differs from Scherer's, in that only RNs were sampled, and the setting was a large, midwestern teaching hospital. In addition, a random sampling technique was used to draw a representative sample from three of the nursing divisions within the institution.

Stilwell (1991) found that 86% of her sample (n = 168) of RNs had fewer than 3 hours of instruction related to the use of restraints on adults. Twenty-nine percent of this sample identified medications as the primary alternative to physical restraint, with sitters (19%) and electronic devices (14%) as additional methods. A disturbing 27% of Stilwell's sample were not aware of alternative methods available within their work settings.

Strumpf and associates (1992) also recognized the need of education to reduce the use of physical restraints. A program aimed at understanding the patients' behavior was developed and included discussions about myths, effects of restraints, legal and ethical issues, change processes, assessment and interventions for the risk of fall, treatment interference, and disruptive behavior. Results of their pilot study showed that while the participants had a significant decrease in their beliefs about restraint efficacy, they were still willing to use restraints. However, they utilized a much broader range of alternative methods aimed at activities and psychosocial care.

METHODS

Instrument

The Attitudes Toward the Use of Restraints (ATUR) (Scherer, 1991) instrument was used to measure nurses' attitudes about restraint use. The ATUR is an 11 -item instrument with a 5-point Likert scale ranging from 1 ("strongly disagree") to 5 ("strongly agree"), with 3 representing "undecided." Higher scores indicate a more positive attitude about use of restraints, a positive attitude being that the nurse would be less likely to use restraints on a confused patient. Scherer reported that content validity was evaluated by a panel of experts. Revisions were made after a pilot study resulted in an instrument with a reliability coefficient of .67 for the 11 items (Scherer, 1991). In the present study, negative responses were reverse coded prior to data entry. Sample questions were, "I feel that family members have the right to refuse the use of restraints" and "It seems that patients become more disoriented after the restraint has been applied." A total mean score was calculated for the ATUR.

Additional data obtained from the nurses included total years of nursing experience and educational level. Information about the unit on which the subject worked was collected in order to classify the data by nursing division - medical, surgical, or critical care divisions. On the post-test 2 measure, a question was added asking for the nurses' source of information about the change in restraint policy. Table 1 summarizes the sources from which nurses received information about the revised policy. Evidently, many of the subjects used multiple sources, with inservices, staff meetings, and the video on the revised restraint policy and procedure being the most common.

The intervention consisted of a hospital-wide effort to communicate both the changes in hospital policy toward the use of physical restraints and specific suggestions about alternative methods to control patient behavior. The policy was distributed to all nursing units and educators, and was the basis for all sources of information listed. In addition to information on documentation and requirements for physician orders, the policy included detailed descriptions of alternatives to restraint use, assessments of patient behavior, possible adverse effects of restraint use, and the need for patient observation and provisions of basic needs during periods of restraint use. The policy also differentiated between and listed devices requiring a physician's order as a restraint and those not requiring an order, since they are considered a regular part of a procedure (for example, arm restraints during mechanical ventilation, which are considered restraints; and bedrails, which are not).

Two videotapes were prepared. One provided extensive information about the legal and ethical aspects of using restraints, the need for adequate documentation, the assessment of the meaning of the patients' behavior, potential underlying medical problems that could cause confusion, components of an environmental assessment, and the nurses' emotional responses to using restraints. The second videotape emphasized adverse physical and psychological reactions of patients, alternatives to physical restraints (diversions, presence of family members, recliner chairs, day rooms, or sitters), and components of hospital policy related to restraint use. While drugs (chemical restraints) were not offered as an alternative to physical restraints, nurses were encouraged to determine if the patient's confused state could be drug-induced and to notify the physician when appropriate.

Table

TABLE 1Sources of Information on Revised Restraint Use Policy (n=1 44)

TABLE 1

Sources of Information on Revised Restraint Use Policy (n=1 44)

Subjects and Selling

A proportional, stratified, random sample was selected from the total population of staff RNs working in the medical, surgical, and critical care divisions of a large, suburban medical center. These divisions are considered to represent separate clinical specialties within nursing. Three hundred fifty-one questionnaires were mailed through interdepartmental mail for the pretest; this represents 48% of the total population in all three divisions. Completion and return of the questionnaires was considered presumed consent for participation in the study. Subjects were instructed not to put their names on the instrument. Social security numbers were included on the pretests, and the post-tests were mailed 10 weeks later to all pretest respondents. The 10-week time frame was necessary to allow for dissemination of information about the policy change.

Once the post-test data were received, the matching sheets with names and social security numbers were destroyed. The 144 questionnaires included in the analysis of pretest/post-test data represent a return rate of 41% of the questionnaires mailed for the pretest and 62% mailed for the post-test. A final sample of 144 completed questionnaires was available for the longitudinal data analysis. The proportion represented by each division was 24% medical; 30% surgical; and 45% critical care. The attempt to obtain a stratified, random sample achieved mixed results. While the surgical division was overrepresented, the medical and critical care divisions were underrepresented.

The institution where data were collected is located in a primarily middle- to upper-class suburb near a large, midwestern city. A broad range of services are offered, including pediatrics, neonatal intensive care, gerontology, and specialty clinics. The emergency department is ranked as a Level ? trauma center with correspondingly high-acuity intensive care units. Nurses who work in this institution are generally well educated, with approximately 41% of the sample holding bachelor's degrees in nursing. The profile of a typical subject is a 35-year-old with a BSN degree, and 10 years' experience as an RN. The nursing units in this hospital are grouped into four divisions, specifically, medical, surgical, mother-baby, and critical care. Only the three divisions in which restraints are typically utilized were included in the sample.

Table

TABLE 2Summary of Descriptive Statistics of Pretest and Posttest Date (n = 144)

TABLE 2

Summary of Descriptive Statistics of Pretest and Posttest Date (n = 144)

RESULTS

Descriptive statistics were used to summarize the pretest and post-test data (Table 2). Findings from pretest data were: responses to the questions asking who has the right to refuse the application of restraints indicate that nurses were undecided about nurses' and patients' rights, but relatively clear on the issue of the families' rights. The greatest variation on the pretest data (SD= 1.19; mean = 2.4) regards the item asking subjects whether families should have the right to refuse /resist the application of restraints on a family member. Clearly, the majority of the nurses did not feel that families can refuse the application of restraints on a confused family member; 68% percent of the subjects responded "disagree" or "undecided" on this item, with responses ranging from 1.21 to 3.59. In contrast, subjects were not as clear on whether or not they, themselves, (mean = 3.3; SD = 1 .09), or the patients (mean = 3.1; SD = 1.13) could refuse the application of restraints.

The subjects were fairly consistent in their feelings about the reasons for using restraints. The greatest agreement (mean = 4. 6; SD = .64) concerned the question of whether or not restraints are a form of punishment for the patient. Because of the reverse coding, a higher score indicates a more positive attitude about restraints (restraints were not viewed as a form of punishment, which is interpreted as a positive attitude). The item asking whether staff used restraints because they were short-staffed was also reverse coded. The mean (4.3; SD = .92) shows the subjects' congruence in their opinions that staffing problems are not adequate justification for using restraints. However, the nurses felt that hospitals do have the obligation to use restraints if a patient is in danger of being injured (mean =3. 7; SD = 1.08).

Subjects were fairly consistent in their feelings about the necessity of using restraints on a patient. Nurses were in agreement that they felt bad when a patient became more upset after applying restraints (mean = 3.7; SD = .94). Although the nurses were embarrassed when family members entered the room of a restrained patient (mean = 4.1; SD = .81), they did feel comfortable taking care of a patient who was in restraints (mean = 3.8; SD = .93). Even though patients in restraints were confused and may not be aware of their surroundings, the subjects still felt it was very important that they showed care toward the patients (RN care, mean = 4.5; SD = .68). The literature indicates that often patients become more disoriented after restraints are applied; however, in this sample, the subjects did not think increased confusion was a problem after the application of restraints (mean = 2. 4; SD = .94).

Initially, an ANOVA was done to determine if the nurses from the three divisions were similar in their preexisting scores (pretest) toward use of restraints. There were no significant differences between the medical, surgical, or critical care nurses' attitudes toward restraint use on the pretest scores. The next step in the analysis procedure was to calculate a repeated-measure ANOVA (Table 3), comparing the three divisions on the pretest and post-test scores to detect any change in attitude resulting from the revision in hospital policy. The analysis showed a main effect for the nursing division (p = .02) but no treatment (change in policy) or interaction effect. The post hoc Scheffe's multiple-comparison test indicated that the divisions that were significantly different from each other on their attitudes toward restraint use were the medical (mean = 3.86) and critical care (mean = 3.63) divisions. The implementation of the policy did not significantly change nurses' attitudes about restraint use.

However, it does seem that nurses working in the medical and critical care divisions have significantly different attitudes, which are perhaps related to their clinical specialty, with medical-division nurses less inclined to use restraints. It should be noted that the mean score across the 11 items (3.6) indicates that nurses were relatively undecided about using restraints and that these perceptions did not change as a result of a change in policy (Table 3).

DISCUSSION

The nurses' perception that family members did not have the right to refuse the application of restraints on a confused patient supports Scherer and associates' findings (1991). However, the nurses in the present study were less certain about whether or not patients should have the right to refuse the application of restraints (pretest data). The sample size and educational level of the nurses in this study may explain this uncertainty. Scherer had only 17 RNs, of which only 12% held baccalaureate degrees, whereas in the sample for the present study, 41.2% (n = 93) had baccalaureate degrees. In addition. Scherer and associates collected data in a nursing home facility, whereas a major medical center was the data collection site in the present study. The responses on the remainder of the items were similar to those of Scherer and associates. The two samples show agreement that restraints were not a form of punishment for patients, that nurses felt comfortable taking care of restrained patients, and that nurses felt bad if patients became more confused after the application of restraints. Neither group felt that confusion increased after the application of restraints, even though the literature suggests that confusion is often an outcome of restraint use.

Table

TABLE 3Comparison of Pretest and Posttest Scores of RNs Using Repeated Measures ANOVA

TABLE 3

Comparison of Pretest and Posttest Scores of RNs Using Repeated Measures ANOVA

Eighty-three percent of Ramprogus and Gibson's (1991) subjects identified injury to the patient as a negative consequence of restraint use, and Werner and associates (1989) documented increased agitation in patients after restraints were applied. In addition, 62% of Werner's subjects experienced a fall while restrained. Tinetti and associates (1992) also documented significantly more falls in restrained patients (17% versus 5% for unrestrained).

Examination of the data shows that the nurses from the three divisions are not significantly different on the pretest measure. Nurses across the medical, surgical, and critical care divisions shared essentially the same views about the application and the use of restraints before the implementation of a new restraint policy. After the dissemination of information about the revised policy through multiple sources, the attitudes of the nurses in the medical division became significantly more positive (they seemed less likely to use restraints). Scherer and associates (1991) compared RNs, licensed practical nurses, and nursing assistants and found no differences between the groups in their attitudes toward restraint use. The fact that the nurses in the medical division had a significant change in their attitude may indicate that the transmission of information about the revised policy within the medical division differed from the other areas. However, as Table 1 indicates, staff meetings and inservices were the most common sources of information about the new policy. Perhaps nurse managers or educators in the medical division emphasized the changes in restraint-use policy and reinforced the use of alternate restraint methods.

Clinical Implications

Clearly, the use of restraints has mixed consequences and may depend on the clinical setting. The trend toward a restraint-free environment is laudable; however, it does not seem appropriate for all confused patients. Changing the policy was beneficial in changing some attitudes about restraint use, but a more comprehensive program may be necessary before nurses' practice also changes. A major shift in how nurses conceptualize confusion will have to occur if practice is to change. From the very beginning of their education, students should be taught to evaluate the causes of patients' behavior, not merely the behavior itself. Focus should center on providing diversional activities, environments that facilitate safe ambulation for the elderly, and initial assessment of and support for independence in activities of daily living.

It can be very difficult to change the attitudes of nurses who have been in practice for many years. A multimedia approach, utilizing role play or experiential activities, may be necessary to emphasize the patients' perspective on being restrained. In addition, it is important to realize that experienced nurses may feel vulnerable as they make the transition to a more restraintfree patient environment. Future studies should not only examine changes in nurses' attitudes but also determine whether the type of restraint utilized also changes. While alternatives to physical restraints can be found in the literature (Strumpf, 1990), future research should evaluate the efficacy of these alternatives.

REFERENCES

  • Coberg, A., Lynch, D., Mavretish, B. Harnessing ideas to release restraints. Geriatric Nursing 1991; 12(3): 133-134.
  • Macpherson, D., Lofgren, R., Granieri, R., Myllenbeck, S. Deciding to restrain medical patients. I Am Geriatr Soc 1990; 38:516-520.
  • Ramprogus, V., Gibson, J. Assessing restraintsNursing Times 1991; 26(87):45-47.
  • Scherer, Y, Janelli, L., Kanski, G., Neary, M., North, N. The nursing dilemma of restraints. Journal of Gerontological Nursing 1991; 17(2):14-17.
  • Stilwell, E. Nurses' education related to the use of restraints. Journal of Gerontological Nursing 1991; 17(2):23-26.
  • Strumpt, N-, Evan, L., Wagner, ]., Patterson, J. Reducing physical restraints: developing an educational program, journal of Gerontolagica! Nursing 1992; 18(11):21-27.
  • Tinetti,M., Liu, W., Ginter, S. Mechanical restraint use and fall-related injuries among residents of skilled nursing facilities. Ann Intern Med 1992; 116:369-374.
  • Werner, P., Cohen-Mansfield, J., Braun, J., Marx, M. Physical restraints and agitation in nursing home residents. J Am Geriatr Soc 1989; 37:1122-1126.

TABLE 1

Sources of Information on Revised Restraint Use Policy (n=1 44)

TABLE 2

Summary of Descriptive Statistics of Pretest and Posttest Date (n = 144)

TABLE 3

Comparison of Pretest and Posttest Scores of RNs Using Repeated Measures ANOVA

10.3928/0098-9134-19950201-09

Sign up to receive

Journal E-contents