Study setting and population
A cross-sectional study was conducted among nurses working at the National Institute of Mental Health in Sri Lanka from January to June 2018. National Institute of Mental Health is the largest tertiary care hospital in Sri Lanka intends to achieve the mental health needs of the community. Institute consists of many specialized units in order to achieve mental health goals of its’ residents such as a psychiatric intensive care unit, perinatal psychiatric unit, forensic psychiatry unit, adolescent psychiatric unit and learning disability units. The study was conducted at the National Institute of Mental Health (NIMH) as it is the center of excellence which provides mental health care in Sri Lanka.
The sample size was calculated using the equation of n = z2 p (1-p) / d2 where n = sample size, z = 1.96; critical value of specified confidence at 95% confidence interval. This is the first study of physical restraining among nurses in Sri Lanka and there is no previous study to help estimate (P). In such a situation, the authors of this study recommended that ‘n’ should be calculated using P = 0.5 (50%) and the minimal sample size was calculated as 267. The sample was further increased by 20% to account for contingencies such as non-response or recording error. Therefore, the final sample size was 308. Registered nurses working in NIMH were included while nurses on maternity leave, participating in in-service education programs during the data collection period and the nurses who didn’t give the informed written consent were excluded from the study.
Data collection and data collection tools
Study participants were selected by non-probability convenience sampling method. Data collection was done using a pre-tested, self-administered questionnaire which contained both open and close-ended questions. A pilot study was conducted using the same questionnaire for ten nursing officers in a psychiatric unit in a general hospital in Sri Lanka to assess the feasibility of the study and drawbacks of the questionnaire.
Forty-seven questions under four sections were created to achieve the aim of the study. First part consisted demographic characteristics, current working place, education qualification and in-service education regarding physical restraining. Second part contained thirteen questions to assess knowledge regarding physical restraining. A total of 13 questions were asked to assess knowledge regarding physical restraining. A scoring system was used to analyze responses to closed ended questions on knowledge. Every correct response was given score of “1” and incorrect response was given “0”. Any nursing officer who did not answered was considered as an incorrect response. Knowledge score has been categorized in to two groups by considering the median of the values.
Third part consisted of fourteen questions to assess nurses’ attitudes regarding physical restraints. Response to the attitude was measured on a 4-point Likert scale. Participants were asked to respond each of the items on whether they ‘Strongly agree’, ‘Agree’, ‘Disagree’ or ‘Strongly disagree’. Each item was given a score of “4” for ‘strongly agree’ to “1” for ‘Strongly disagree’ and vice versa for negatively phrased items. Items 3, 4, 9 and 10 were negative items and their scores were reversed.
fourth part included fourteen questions on practices regarding physical restraints. Practices such as alternative nursing measures, involving experienced nurses, explanation about restraint to the patients and caregivers, expression of warning signs and check the patient during the period of restraint and after were assessed under the fourth session. Most of the questions were reflective towards more favorable practices. Each question was given a score of “3” for ‘Always’ to “1” for ‘Never’ having adopted such practices and the negative items was reverse scored. Item 8 and 9 were negative items and needed to be reverse scored.
The questionnaire was initially developed in English and then back translations were done into Sinhala and Tamil which are the native languages of Sri Lankans.
Data was collected after obtaining written informed consent from nursing officers after giving a clear explanation of the study to the nursing officers.
The questionnaire was distributed among the nursing officers after the objectives and the procedure of the study were explained. Each participant was informed they could withdraw from the study at any time. And also information sheets and volunteer consent forms were handed over to the sample prior to the data collection to give enough time to understand the information sheet. Informed written consent was obtained by a volunteer consent form from each participant after clear explanation of the purpose of the study. The information sheet was developed according to recommended format.
Data was collected within a minimum time period without a contaminating with the other wards. The decision to participate in the research was decided by themselves without any influence of the investigators. The participants completed the questionnaires without any interruption to their duties.
Confidentiality of the research was ensured at all stages of the study. The names of participants were not recorded on the questionnaires and an index number was given to each participant for maintaining the confidentiality. The information collected from this research project kept confidential. All the collected data were stored in a password protected electronic device and it will be permanently deleted after 5 years. Internal consistency reliability for each part of the questionnaire was highly reliable (Cronbach’s alpha was > 0.90).
Privacy and confidentiality were ensured at all stages of the research. The names of participants were not recorded on the questionnaires and an index number was given to each participant for maintaining privacy and confidentiality. The questionnaire didn’t contain any sensitive questions which could psychological trauma or embarrassment to the participants in any way. Contributors were given the opportunity to ask questions and register any complains via contacting the investigators and the supervisor through giving contact details. Moreover, participants were informed that they could withdraw from the study at any time without any consequence to their duties.
Confidentiality of the database was maintained under a completely secured way and all data were handled only by the investigators of the study. Data and other information weren’t given to any third parties. All collected data will be stored in an electronic device for 5 years under password protected and they will be permanently deleted. Hard copies will be kept under lock and key for 5 years and they will be burned according to international norms.
Statistical analysis
All the questionnaires were numbered accordingly and entered into the Excel worksheet. All the multiple data were entered according to the cord system. Data was analyzed by Statistical Package for Social Sciences (SPSS Version 23). Descriptive statistics were performed on the responses to the knowledge, attitudes and practices individually.
A total of 13 questions were asked to assess knowledge regarding physical restraining. The chi - square was used to assess knowledge with three demographic characteristics; working experience, whether to get a formal education about physical restrain and the working place.
Participants were asked to respond each of the items on whether they ‘, ‘Agree’, ‘Disagree’ or ‘Strongly disagree’. Each item was given a score of “4” for ‘strongly agree’ to “1” for Response to the attitude was measured on a 4-point Likert scale with numbers referring to 1 = Strongly disagree, 2 = Disagree, 3 = Agree and 4 = Strongly agree. And vice versa for negatively phrased items.
Most of the questions were reflective towards more favorable practices. Each question was given a score of “3” for ‘Always’ to “1” for ‘Never’ having adopted such practices and the negative items was reverse scored. Then the level of practice was cross tabulated against the selected variables. Cronbach’s alpha was used to estimate the reliability of a psychometric test in statistic.