The present non-randomized controlled clinical trial was conducted in the emergency department at selected hospital affiliated to Kermanshah University of Medical Sciences, Iran. The research units included newly admitted Level-3 patients (based on the Emergency Severity Index (ESI) triage), who met the inclusion criteria. Sample size was calculated 32 for each group using previous studies (5) based on the equation of (type I error: 1.96, type II error: 0.85, test power: 0.80, and observed effect size: 0.70). Considering the 10% dropout, the final sample size was considered to be 35 patients. Therefore, 70 patients were selected by convenience sampling method and assigned into intervention and control groups.
Because the presence of a caregiver for a number of patients may cause discontent to patients without a caregiver that is ethically incorrect, there was no possibility of random assignment. The intervention group members were selected in even days and control group members in odd days. It was impossible to meet the two groups each other according to the determination of the status of the patients within less than six hours (discharge, hospitalization, referral or dispatch to another center), which is important indicator of emergency department.
Inclusion criteria were the age of 18 to 65 years, consciousness, communication and cooperation, being subjected to invasive nursing procedures while admission in the emergency room, presence of a caregiver, no anxiety and psychological disorders, no history of diabetes, no use of hypnotics and tranquilizers, no addictive and psychotropic drugs, no severe pain, and no taking sedatives on admission. Exclusion criteria included any complication during the intervention (failure of invasive nursing procedure or the alteration of patient triage to lower level) and unwillingness to continue to participate in the study.
Data were collected using demographic and clinical information questionnaire, Spielberger State-Trait Anxiety Inventory (STAI) and Visual Analogue Scale (VAS). The demographic and clinical information questionnaire consisted of age, gender, marital status, living status, educational level, employment status, type of invasive nursing procedure and history of doing this procedure and family member relationship with the patient (in the intervention group). Content validity was used to determine the validity of the demographic and clinical information questionnaire. Thus, the relevant form was prepared after studying the relevant references, and corrected with the comments of ten faculty members of the Nursing and Midwifery School at Shahid Beheshti University.
The STAI is a standardized and commonly used questionnaire developed by Spielberger et al. in 1970 with two parts of state and trait anxiety. The state anxiety scale measures a person's feelings at the time of answering (5), including 20 items, each rated on a 4-point Likert scale. The total score ranged from 20 to 80 and higher scores indicated higher anxiety. The trait anxiety indicates mild (the score of 21–40), moderate (the score of41-60) and severe (the score of 61–80) anxiety. In a study of Najafi et al. (2014), the reliability coefficient of this tool with Cronbach's alpha was 0.89. Its concurrent validity has been confirmed by clinical interviews and a number of anxiety assessment tools . In this study, the reliability was determined using Cronbach's alpha test which was 0.90, indicating high and acceptable range. The VAS is the most widely used (one-dimensional) tool for measuring pain around the world ) using a straight 10-cm line representing the words “no pain” at the left-most end and “worst pain imaginable” at the right-most end. The patients mark this line according to the extent of their pain, and then a numerical representation of their pain severity is determined by measuring the distance between the left-most end and the point where the patients have marked . This tool has been widely used in national and international studies and its validity and reliability have been demonstrated .
After obtaining the necessary permits and coordinating, the researcher went to the selected hospital affiliated to Kermanshah University of Medical Sciences, introduced himself to the research units, provided sufficient explanations to them and obtained informed written consent if they wish to participate in the study. At first, the demographic and clinical information questionnaires were completed for all the research units on admission, and then the STAI questionnaire was filled in for both intervention and control groups. Next, in the intervention group, an invasive nursing procedure recorded in the emergency record, including peripheral intravenous cannulation, intramuscular injection and venipuncture, was performed by the nurse in the presence of a family member. The family member (spouse, father, mother, child, sister, brother, or person who had a kinship or emotional relationship with the patient) was selected with the patients' willingness to attend the emergency room monitored by them until hospitalization. The caregiver-patient relationship was recorded in a demographic and clinical information questionnaire. Necessary explanations were given to the caregiver in the intervention group before the study, who was asked to encourage the patient during the invasive procedure and to provide physical and psychological support. In addition, compliance with the requirements such as non-intervention in the process of invasive procedure was taught. After rubbing alcohol, NOKAST pink peripheral venous catheter  was used for venipuncture and a 3-cc syringe with 23G needle (Avapezeshk Company) was applied for intramuscular injection and blood sampling.
In the control group, the invasive nursing procedure recorded in the patient's file was performed without family presence. In fact, family members of the control group were encouraged to leave their patients in different ways, such as explaining the aims of the study. Immediately after the end of the invasive nursing procedure, the pain intensity was measured by VAS and the STAI questionnaire was re-completed in two groups. It should be noted that all medication and treatment were tailored to the patient's condition based on the routine emergency department for both groups. It was impossible to adjust to ambient conditions that affect the patient's level of anxiety (such as light and sound) due to the nature of the emergency department, and it was impossible to blind to the nature of the research.