This study is a randomized controlled trial that will compare two arms: an experimental intervention group that receives a decision aid consisting of an educational video and targeted manual providing a protocol for the family members' disclosing terminal illness to the patients and aimed at improving communication, satisfaction with decision making process, and quality of life between patients and their family versus an attention control group that receives a video and non-tailored manual on pain control.

Inclusion and exclusion criteria:

This study is aimed at decreasing the decisional conflict among family members who are ready or not to disclose patients' terminal illness. The decision aid will lead to effective communication between patients and family and the effective decision making process will lead to the final outcome to improve the quality of life in patients and their family. This trial is limited to situations where the family and their patients are cognitively capable of participating in decision-making process.

Measures:

Baseline Measures. At baseline, before randomization, self-reported questionnaire is used to collect data. To correct possible confounders in this study, family' background variables are assessed on entry to the study, including sociodemographic and decisional conflicts, decisional-role preference, stage of readiness, level of anxiety and depression, and QOL.

Family decisional conflict is assessed using four subscales of the Decisional Conflict Scale which assessed how well informed participants feel about their choices and the associated benefits and risks; the clarity of their values; the support they have in the decision-making process; and their level of uncertainty. It consists of 16 items followed by a Likert response of 1, strongly agree, to 5, strongly disagree. A mean score is obtained for each participant. The Decisional Conflict Scale (DCS) is translated into Korean by two bilingual individuals, using a forward-backward translation procedure, and completed pilot test.

Family preference for role in decision making is assessed using a scale modeled on an instrument developed by Degner et al. We make family/patient dimension from modifying Degner and Sloan's second Control Preference Scale set, family/physician dimension. The options of modified set of five cards (family/patient dimension) range from the family making the decision mainly, through a collaborative model where the family and patient jointly decided, to a scenario where the patient made decision mainly. Family is asked to indicate whether they preferred an independent role (i.e., "that the family makes the final decision regarding which treatment the patient will receive or after considering patient's opinion"), shared role (i.e., "that the family and patient together select which treatment is best for the patient"), or a dependent role (i.e., "that the patient alone makes all the decisions regarding the patient's treatment or considers the family's opinion"). Family participant chose the decision-role statement that best described his/her preferred role in decision making. The Control Preference Scale (CPS) is translated into Korean by two bilingual individuals, using a forward-backward translation procedure, and completed pilot test.

Levels of family anxiety and depression is assessed by the Hospital Anxiety and Depression Scale which consists of 14 items, seven on anxiety and seven on depression, forming two sub-scales. Each item has four descriptive response options to be scored on a 0 to 3 scale. A value of 0 corresponds to not having the symptom, 3 corresponds to having the symptom to a high level. Scores for each of the two sub-scales are constructed by simple summation of its seven items. The developers have suggested that sub-scale scores of 0-7 represent non-cases, 8-10 represent doubtful cases, and 11-21 represent definite cases.

The Baseline Patient's Generic QOL:. Because we expect the effect of intervention for family to accompany change of patients' outcome, we measure patients' QOL and quality cancer care at baseline to observe secondary outcome related with effect of the decision aid.

Patient's QOL is assessed with QCQ-EOL. To correct possible confounders in this study, patient' background variables are obtained on entry to the study including sociodemographic and medical records.

Outcome Measures:

The primary outcome to be considered is the potential efficacy of this aid including family's decisional conflicts, satisfaction with decision making, and decisional-role preference, decisional regret, and physician and family acceptance of the decision aid.

Family satisfaction with decision making is assessed using the effective decision-making subscale of the decisional conflict instrument [O'Connor AM. Validation of a decisional conflict scale. Med Decis Making. 1995 Jan-Mar;15(1):25-30]. The subscale was four items regarding satisfaction with decision making (being informed, reflection of participant's own values, likelihood to comply, and satisfied). Scores for family satisfaction are reversed so that the higher scores reflect higher level of satisfaction (ie, a reversed Likert scale in which 5 is the highest scored). A mean score is obtained for each patient.

Physician and participant's acceptance of the decision aid is addressed using two questionnaires followed by a Likert response of 1, strongly agree, to 5, strongly disagree (i.e., family assessment item, "The decision support was useful to me in reaching a decision", "I would recommend that others facing the decision use the support", "The decision support interfered with my relationship with patient", Physician assessment item, "I feel the family has made an informed choice", "The decision support was useful in helping this family understand the risk and benefits of the disclosure on terminal illness", "The decision support did not interfere with my anticipated physician-patient relationship with my patient")[Brundage et al2001]. Decision regret is measured by the Decision Regret Scale which is useful tool for measuring regret after decisions at a particular point in time developed by O'Connor et al. It consists of 5 items followed by a Likert response of 1, strongly agree, to 5, strongly disagree. The definition of decision regret is remorse or distress over a decision[O'Connor]. These items were whether regret was felt to be present, whether the decision was the right one for him/her, whether he/she would make same decision again in the same circumstances, whether the decision was a wise one, and whether the decision had resulted in harm. The Decision Regret Scale is translated into Korean by two bilingual individuals, using a forward-backward translation procedure, and completed pilot test.

Secondary outcomes include the levels of family anxiety and depression,and family's QOL.

Levels of family anxiety and depression is assessed with the Hospital Anxiety Depression Scale. The actual disclosure following use of the material is assessed by following items including chosen disclosure and if terminal illness disclosed "who, and when disclose patients' terminal condition" and it is assessed at 1, 3, and 6 months after sending the material. Levels of family anxiety and depression, family decision regret, family preference for role in decision making, and patient's QOL are assessed at 1, 3, and 6 months after sending the video.

Inclusion Criteria:

Patient:

• age≥ 18years
• cancer patient
• terminally ill
• mentally competent

Family members:

• age≥ 18years
• having terminally ill cancer patients
• Spouse, daughter or son, parent, brother or sister
• having ability to understand the content of educational material

Exclusion Criteria:

• serious comorbidity(stroke, heart attack, acute or chronic respiratory, acute or chronic renal disease, or cancer)
• inability to speak, hear or read Korean
• mentally incompetence (schizophrenia, bipolar, psychiatric or addictive disorder, or severe emotional distress)
• Pregnancy