Advance Directives and End-of-Life Decision Making

Abstract

This article describes critical care nurses’ knowledge, attitudes, confidence, and experiences regarding advance directives and end-of-life decision making.Intensive care units (ICUs) are the site of much end-of-life decision making. Decision making in the ICU should be governed by patients’ wishes, and advance directives are one way patients can make their end-of-life decisions known. Advance directives are defined as mechanisms by which individuals make known how they want medical treatment decisions made when they can no longer make the decisions for themselves. Advance directives can take the form of living wills, healthcare proxies, do-not-resuscitate orders, and durable powers of attorney. Healthcare providers play an important role in patients’ understanding and completion of advance directives. Ideally, advance directives should be completed before an acute illness occurs, in a less hectic and stressful environment. If patients come to the ICU without advance directives, critical care nurses, who spend more time with ICU patients than any other provider does, are in the best position to educate competent patients about advance directives and facilitate the completion of such directives. Providers’ knowledge of and attitudes toward advance directives can be important aspects that influence the effectiveness of the providers’ role in helping patients complete advance directives and in ensuring that patients’ end-of-life wishes are carried out. Helping patients’ family members understand and cope with end-of-life decision making in the ICU environment is likewise important.The purpose of this study was to describe critical care nurses’ knowledge, attitudes, confidence, and experiences regarding advance directives and end-of-life decision making. The aims of the study were to determineEach of the United States has laws/statutes governing advance care planning in the form of directives. New York State, where this study took place, has a healthcare proxy law and a do-not-resuscitate statute. Living wills are legal in New York State on the basis of case law. The federal government enacted the PSDA in 1991; the act mandates that any healthcare institution that receives Medicaid or Medicare funds must inform its patients about the patients’ right to self-determine care at the end of life according to the laws of the state in which the institution resides. Despite more than a decade of laws governing patients’ rights to end-of-life decisions, it is estimated that less than 20% of the US population has completed an advance directive.1–3Few data are available on nurses’ knowledge and skills in advance directives. Crego and Lipp4 found that a volunteer sample of nurses (n = 399) from a midwestern acute care teaching hospital had limited knowledge of advance directives. A 44-item questionnaire developed by the researchers4 was used to assess the nurses’ knowledge. The range of scores was wide (40%–95%), indicating marked deficits for some nurses. Also, more than half the nurses indicated that they did not have a good understanding of advance directives. The nurses surveyed concurred, however, that discussion of advance directives is within nurses’ professional role. A total of 67% thought that a nurse was the most likely and most appropriate care provider to begin discussion of advance planning and end-of-life decisions with patients.Similarly, in a small survey of hospital nurses (n = 112), Wood and DelPapa5 found that although nurses had favorable attitudes toward advance directives, the nurses’ knowledge was inadequate. A total of 76% of the nurses had low knowledge scores on questions on laws about advance directives.Solomon et al6 explored nurses’ and physicians’ knowledge of and attitudes toward national recommendations regarding patients’ rights to forgo life-sustaining treatments. The investigators concluded that changes in the care of dying patients have not kept pace with national regulations, in part, because physicians and nurses disagreed with and were unaware of some key guidelines.Critical care nurses are often involved with patients and patients’ families in end-of-life decision making.7 The nature of the critical care environment alone makes addressing advance directives unique compared with other aspects of healthcare. First, the families of most patients are experiencing extremely high levels of stress because of the severity of their loved ones’ illnesses and the uncertainty associated with the outcome.8 Second, the use of sophisticated interventions and equipment such as ventilators and ventricular assist devices make it difficult for patients and their families to understand the scope of the interventions that may be used.9,10 This lack of understanding leads to increased anxiety and further complicates making decisions about advance directives.11 Finally, patients often cannot participate in the advance directive process, further exacerbating the difficulty of the decision making.12An essential contribution of critical care nurses who work with patients and patients’ families is interpreting the patients’ experience of illness and treatment to assist the families in decision making when the patients are unable to make decisions.13 In order to interpret patients’ experience of illness and treatment and assist patients’ families in making decisions, nurses must maintain open lines of communication with patients and the patients’ families. For example, critical care nurses’ interactions with patients and/or patients’ families who were signing do-not-resuscitate consents have been investigated by using a qualitative perspective (grounded theory).14–16 The findings of the studies indicate the importance of establishing a trusting relationship with patients and the patients’ families, maintaining open communication with the patients and families, and serving as the patients’ advocate. Other important roles identified by nurses include acting as a decision maker and educating patients and the patients’ family members.15,17 In a qualitative study of family decision making for ICU patients, Jacob17 found that skillful and supportive involvement of care providers was related to more positive long-term outcomes for patients’ family members. However, little empirical evidence is available on critical care nurses’ knowledge of and attitudes toward advance directives or on the roles the nurses are actually performing.Baggs and Schmitt11 assessed the current research on end-of-life decisions in adult intensive care. They concluded that few studies are available on nurses’ involvement in end-of-life decision making. ICU nurses reported frustration about their limited role in this decision making, and ICU nurses disagreed and were confused about the best way to care for patients at the end of life. Additional data on nurses’ knowledge of, attitudes toward, and experiences with advance directives and end-of-life decision making must be collected and analyzed before an intervention can be implemented to inform critical care nurses about the best way to assist patients and patients’ families with advance directives.A survey design was used to conduct a descriptive, correlational study of a random sample of nurses currently practicing in critical care.A random sample of members of the American Association of Critical-Care Nurses (AACN) in New York State who are registered nurses and who described their work status as full-time or part-time on their AACN membership application received the mailed survey. A state survey rather than a national survey was used because we wanted to measure critical care nurses’ knowledge of the state laws and statutes governing completion of advance directives in the state where the nurses lived and because the logistics of scoring the knowledge subscale for all 50 states would have been prohibitive. Using power analysis, we determined that a minimum of 107 subjects would have 80% power to detect a medium effect size of 15% at α = .05 for the multiple-correlation question of whether relationships exist between (1) selected demographic characteristics and (2) knowledge, attitude, confidence, and experience regarding advance directives and end-of-life decision making.Because of the large number of items on the survey and the expected low response rate of mail surveys, the AACN membership was oversampled. The AACN membership in New York State is 4876 nurses; the names of a random sample of 1000 (20.5%) members from New York State were obtained from AACN. The return rate for the survey was 21%; a total of 210 surveys were usable for data analysis.The Knowledge, Attitudinal, Experiential Survey on Advance Directives (KAESAD) instrument was developed by Jezewski et al18 to measure respondents’ knowledge of, attitudes toward, and experiences with advance directives and end-of-life decisions. Reliability and validity of the instrument were established through an expert panel (n = 7) and a test-retest pilot study with 56 oncology nurses. The panel consisted of experts in end-of-life care and advance directives and represented the disciplines of nursing, medicine, law, and bioethics. The panel provided feedback about each of the 110 items in the original draft of the survey. On the basis of the opinions of the panel and the judgment of the investigators, changes were made to 22 items, 9 items were added, and 4 items were deleted.After the changes were made, the test-retest pilot study with the 56 oncology nurses was done. A convenience sample of 18 graduate nursing students was also used to establish test-retest reliability. Test-retest reliability and the Cronbach α were analyzed. The results of the analysis indicated test-retest scores for the various scales (r = 0.51–0.90), test-retest proportion of agreement for individual items (0.71–1.0), and internal consistency for the various scales (Cronbach α .58 to .95).The KAESAD instrument used in this survey consists of 115 items. Table 1 gives the principal components of the survey, the numbers of items in each section, and the internal consistency (Cronbach α) of the attitude, confidence, and experience subscales for the data from the pilot study.Of the 115 items, 85 are divided into 5 scales designated to measure knowledge, attitudes, confidence, and experience regarding advance directives and experience with end-of-life decision making. The first of the 5 scales encompasses 30 items that contain questions about critical care nurses’ knowledge of advance directives. The 30-item scale comprises 3 subscales related to general knowledge of advance directives, the PSDA, and state laws governing advance directives. Scores for each of 3 subscales and a total knowledge score were calculated. Respondents were asked to respond to each question by checking yes, no, or don’t know. The don’t know choice was provided to minimize guessing the correct answer.The second scale consisted of 20 items about nurses’ attitudes toward completing advance directives and end-of-life decision making. Nurses were asked to respond to each item by using a 6-point Likert scale (1= strongly disagree, 6 = strongly agree). Items addressed nurses’ attitudes toward patients’ rights, starting and stopping life support, artificial hydration and nutrition, nurses’ role in informing patients, and assisted suicide.The third scale consisted of 5 items to assess nurses’ clinical experience with advance directives. Nurses were instructed to answer yes or no to the items. Two additional items not included in the scale asked nurses (1) Have you provided treatment to patients whose advance directive indicated otherwise? and (2) Have you observed others providing treatment to patients whose advance directive indicated otherwise? Nurses responded yes or no. These questions were separated from the third scale in order to improve the internal consistency of the scale.The fourth scale consisted of 11 items on nurses’ confidence in helping patients complete advance directives. Items addressed the nurses’ confidence in answering the questions of patients and patients’ families, teaching others about advance directives, mediating in disagreements related to advance directives, and advocating for patients’ advance directives. Nurses responded by using a 5-point Likert scale (1 = not at all confident to 5 = very confident).The fifth scale consisted of a series of 20 statements about the nurses’ professional experience with end-of-life decision making. Again the nurses were asked to respond to each item by using a 6-point Likert Scale (1 = strongly disagree to 6 = strongly agree). Items included statements about professional opinions related to communicating prognosis, how patients make decisions, whether the presence of advance directives encourages communication, and nurses’ roles.The instrument also contained 26 items on demographic characteristics. The responses to these items were used to assess personal, professional, and institutional variables. Personal variables included race, religion, sex, marital status, age, and whether the respondent had completed an advance directive for himself or herself. The professional variables included professional education preparation, clinical practice site, current position, AACN certification status, work status, and amount of formal instruction on advance directives. Institutional variables included geographic location of the workplace, the presence of an ethics committee in the work-place, and how patients’ advance directives are formally communicated within the workplace.The last item of the instrument was an open-ended question: What do critical care nurses need most to increase their ability to assist patients with advance directives? The written responses to this question are being analyzed by using qualitative techniques and will be reported separately.The 1000 nurses were mailed the KAESAD survey. Those nurses who might have discarded their survey were given a telephone number on a separate sheet of paper to request a new survey if they desired. This strategy was selected, rather than a second letter only to nonrespondents, to ensure anonymity of respondents. The survey was mailed by using a bulk mail permit and the university mailing service. A stamped self-addressed envelope was included to facilitate return of the survey. Approval to conduct the study was obtained from the University at Buffalo institutional review board.Data were entered into Excel (Microsoft Corp, Redmond, Wash) and imported into SPSS, version 10.0 (SPSS Inc, Chicago, Ill). Those questionnaires with small amounts of less than 20% items knowledge, attitude, or experience were included in the analysis with the data by the was used for the was at questions were or don’t know. In order to knowledge don’t know were as attitudes toward completing advance directives and experience with end-of-life decision making were to The 3 some of agreement were used to the 3 some of disagree, strongly were used to the Cronbach α for the scale and professional experience with end-of-life decision making were items for these scales were at not as for experience with advance directives were on yes a score of scores were on total scores for the 11 items in the scale and were on a 5-point Likert scale (1 = not at all confident to 5 = very Scores range from 11 to relationships nurses’ personal, professional, and institutional characteristics and their knowledge of advance directives, attitudes regarding advance directives, experiences with advance directives, and confidence in helping patients complete advance directives were determined by using or analysis of as appropriate to the of of the were and The had a or were nurses, and one third were had formal instruction in advance directives at their of Respondents were asked to the number of of instruction in advance directives they received their nursing Scores from to 50 A small of the were members of their ethics half of the critical care nurses had an advance directive for and more than half had family with an advance care nurses’ knowledge was by the knowledge scores on 3 subscales and the total knowledge score on the basis of the scores on the 3 Table 3 is an of the and correct scores for the 3 knowledge subscales and the total knowledge first knowledge subscale was used to measure the nurses’ general knowledge of advance directives. The measure included of advance directives and the various of advance directives durable power of healthcare role of proxy decision and legal related to advance directives. scores were for this A analysis of was carried to detect The results indicated that the scores of nurses who were from the scores of nurses who were not critical care nurses (n = than did critical care nurses second knowledge subscale on questions related to the principal components of the Items in this subscale included to inform patients about their rights, and of Respondents on this knowledge third knowledge subscale was used to measure respondents’ knowledge of the laws of New York State on advance directives. on this subscale on the when a can complete an advance the law regarding and and of advance directives from other The respondents’ score was knowledge scores were by using the 30 items from the 3 The total knowledge score was of a 30 or nurses were asked to respond to the series of 20 items regarding the nurses’ attitudes toward advance directives and end-of-life decisions. Table 4 the and of agreement for some of items. of agreement was determined by the and strongly responses into a consistency in this study according to the Cronbach α was (n = to items a role of nurses as for For example, most that nurses should a wishes if the wishes with the nurses’ Also, agreement was high that nurses are for with a if a rights have not been and that nurses should inform patients of their and treatment half of the that some patients to should be made with the that starting or stopping life is the The strongly that patients should the the patients need the may This response the current on at the end of care nurses’ experiences with advance directives were by using a yes or no response to a series of 5 consistency for this scale was Cronbach The yes score was = with an of of the had for patients who had advance directives patients and patients’ families about advance directives and discussion with patients about advance directives Two additional questions were asked about experience with advance directives. half of the of had provided treatment to patients whose advance directives indicated and of had observed others providing treatment to patients whose advance directives indicated treatment was not the as to the included cannot be respondents’ confidence in their role of patients and patients’ families was by using the consistency for this study was for the confidence Cronbach responded by using a 5-point Likert scale (1 = not at all confident to 5 = very The total score for confidence was score = with an of The nurses were confident about the PSDA score and about state laws on advance directives score Respondents were most confident on items with confidence in score and answering patients’ score and patients’ score questions about advance nurses were also asked to respond to 20 statements about professional experience with end-of-life decision making. consistency was Cronbach 5 gives and of agreement of with the for some of the was low for the statements that the amount of time nurses spend advance directives with patients is and that patients have knowledge about their medical and to advance directives that is not time to advance directives and that nurses know the wishes of their patients about advance directives than half that in an advance directive is to the relationship between (1) personal, professional, and institutional characteristics and (2) knowledge, and confidence we found although knowledge scores were with total experience scores (r total confidence scores (r (r and in practice (r knowledge scores did not with of education on advance directives. However, total confidence scores did with of education on advance directives (r of an advance directive by family members was with the respondents’ confidence, and total knowledge Nurses whose family members had an advance directive had scores for experience with advance directives confidence and knowledge than did nurses whose family members did not have advance directives. Also, nurses with advance directives of their had scores for experience with advance directives and confidence than did critical care nurses without advance directives of their education in the on advance directives was with experience and total knowledge Nurses who had education in the on advance directives had knowledge scores and more positive experience with advance directives than did nurses who had no education on advance the of the of this study although any to make a on the basis of the data would be the data do describe various and that at a minimum further Other or and/or further and of the KAESAD instrument be used to about this and critical care nurses. the data do describe and may the about this sample of New York State AACN the nurses were about advance directives in general not about the PSDA or the New York State law on advance directives. certification was with general knowledge The that certification may increase nurses’ understanding of advance directives. The that nurses who had education in the on advance directives had knowledge scores and more positive experiences with advance directives and end-of-life decision making than did nurses who did not have such education may that education in the is an way to make nurses about advance directives and end-of-life decision attitudes a role in for patients’ rights to make decisions about end-of-life care and advance directives. for assisted was of the and most did not have an with stopping treatment treatment had been nurses’ confidence was low regarding the PSDA and New York State law on advance directives, the nurses were very confident in their ability to assist patients and patients’ families in completing advance directives. The high of nurses who provided treatment or observed others providing treatment to patients whose advance directives indicated is have reported that healthcare providers’ of and use of patients’ advance directives are and Schmitt11 that on the need to patients, patients’ families, and care providers in end-of-life decision making, evidence is that physicians often make these decisions with little from In a study by only of do-not-resuscitate were on patients’ et found that was to of their study sample who had advance directives. care nurses in study who had completed their advance directives had confidence scores than did nurses without advance directives, that individuals who complete their advance directives are to assist others in completing such experiences with end-of-life may the hectic pace and high stress levels in an ICU environment. The responses to survey also the in advance directive decision making related to the of care in the This is by the low of agreement to items such as patients have knowledge about their medical and to advance can be in an ICU environment because ICU care is not a of or to have other such as use of and ventricular assist that patients and families are unaware of are and about these patients and patients’ families may not to have need to be to patients and their families in they can way that may improve the of a family members of treatment when advance directives lack to treatment is the family in which family nurses, and can end-of-life care Nurses also need to their to as patients’ making that the patients’ advance directives are and their wishes in this study the of The pilot study indicated that the test-retest reliability for some of the scales was low (r = however, the test-retest proportion of agreement for individual items to was consistency was low in this study for the (Cronbach α = experience with advance directives (Cronbach α = and professional experience with end-of-life decision making (Cronbach α = The of low Cronbach was addressed by at the items not as Additional use and of the KAESAD instrument are for additional of its the although are and may have little influence on the The findings are only of critical care nurses who are members of AACN. The survey was in New York State and may not the knowledge and attitudes of nurses in other The response rate was The of the survey may have to the low response