Research Design
A descriptive cross-sectional study was conducted between April and June 2019. Following permission from the social welfare organisation, the first author contacted nursing home managers and the geriatric ward head nurse to recruit NS providing care for older people. The NS included registered nurses (RNs), ENs, and nursing assistants (NAs).
Setting
In Iran, RNs have at least a four-year bachelor’s degree in nursing science, ENs pass a two-year academic program, and NAs pass a short formal care training course (42). NS provide care for older people with multiple comorbidities, including dementia and physical disabilities, during the day (7 am to 3 pm) or evening/night shift (3 pm to 7 am).
Three RNs (including one head nurse) and two NAs on the day shift and two RNs and two NAs on the evening/night shift provided care to 11 residents in the hospital geriatric ward. Care for older people in Iranian nursing homes involves teamwork, usually managed by a general practitioner. Therefore, staffing patterns differ depending on the residents’ needs. Based on Iranian social welfare guidelines, one RN and one EN on the day shift and two ENs on the evening/night shift were responsible for approximately 50 residents. One NA was assigned to every seven dependent residents (Barthel index 0–49) or ten independent residents (Barthel index 75–99) (43). Weekends have lower staffing.
Participants
All NS, including RNs, ENs, and NAs, with at least one year of experience caring for older adults agreed to participate in the study. Furthermore, they were located in 20 nursing homes and one geriatric hospital ward located in different Tehran areas.
In total, 161 NS, including 43 RNs/ENs and 118 NAs, participated in this study. The participants included 149 NS from nursing homes and 12 from geriatric wards. Participants were 19 to 64 years old (36.7±9.3). Most were women (75.8%), married or cohabiting (51.6%), and had a high school diploma (49.1%). Approximately 42.2% of the participants had less than five years of clinical experience, and 49.7% had less than five years of experience in caring for older people. Participants’ demographics are presented in Table 1.
Table 1
Background data of the nursing staff working in older people care (N = 161)
Gender |
Male | 39 (24.2%) |
Female | 122 (75.8%) |
Age |
Mean ± Sd | 36.7 ± 9.3 |
Clinical Experience |
Less than 5 years | 68 (42.2%) |
5 to 10 years | 45 (28.0%) |
10 to 15 years | 25 (15.5%) |
More than 15 years | 23 (14.3%) |
Clinical experience in caring of older people |
Less than 5 years | 80 (49.7%) |
5 to 10 years | 43 (26.7%) |
10 to 15 years | 22 (13.7%) |
More than 15 years | 16 (9.9%) |
Education |
Finishing elementary school (Cycle in Iran) | 39 (24.2%) |
High school diploma (Completing high school) | 79 (49.1%) |
Degree in nursing (Academic education) | 43 (26.1%) |
Marital status |
Married | 83 (51.6%) |
Never married | 45 (28.0%) |
Divorced | 26 (16.1%) |
Widow | 7 (4.3%) |
INSERT TABLE 1
Data collection
Data were collected using four different questionnaires: (1) sociodemographic, (2) the revised Perceptions of Conscience Questionnaire (PCQ) (8), the revised Stress of Conscience Questionnaire (SCQ) (8), and the Oldenburg Burnout Inventory (OLBI) (44). Appointments were made with each participant to hand in the questionnaires and answer any questions if needed. All distributed questionnaires were collected by the research team, which resulted in no dropout cases.
Instruments
The three published questionnaires were translated from English into Persian by a professional translator. They were then back-translated into English using English-speaking RNs. Qualitative, face, and content validity were determined according to the method described by Polit and Beck (45) by a panel of experts, who assessed the content validity ratio (CVR), item-content validity index (I-CVI), and scale-content validity index-average (S-CVI-Ave). Finally, linguistic validation was performed by the first author and the professional translator to ensure that the phrases of the questionnaires could support the interpretation of the findings.
The sociodemographic questionnaire included age, sex, marital status, education, clinical experience, and care experience of older people.
The revised PCQ consisted of 16 items. NS assessed their PC (its origin, function, and significance) on a 6-point Likert-type scale from ‘1=no, totally disagree’ to ‘6=yes, entirely agree’. According to a factor analysis of the PCQ, Åhlin et al. (8) reported that NS perceived conscience as an authority (items 6,10,14,15), an asset (items 7,8), a warning signal (items 3,4,5), demanding sensitivity (items 1,2,10), or a burden (items 11,12,16) and dependent on culture (items 9,13). Different studies have been conducted using the SCQ (10-12). In our study, Cronbach’s α for the PCQ was estimated to be 0.87. The revised SCQ consists of nine two-part items, including A and B questions that measure stress originating from a troubled conscience based on staff experience in healthcare. The A questions assessed the frequency of certain difficult ethical situations on a 6-point Likert-type scale from ‘0=never’ to ‘5=everyday’. The following B questions estimated the amount of troubled conscience related to each given situation on a 10 cm visual analogue scale from ‘0=no, it gives me no troubled conscience at all’ to ‘5=yes, it gives me a very troubled conscience’. The SCQ index for each item was calculated by multiplying the score from each A question by the score of its analogous B question. The SCQ index ranged from 0 to 25 points for each item. This questionnaire has been used in different countries (24, 25, 46). In the present study, Cronbach’s α for the SCQ was estimated to be 0.82.
The OLBI has two dimensions: exhaustion (affective, physical, and cognitive) and disengagement. Each subscale consists of eight items, four of which are positively and four of which are negatively worded. These were presented in a mixed order with four response alternatives, ranging from ‘1=totally disagree’ to ‘4=totally agree’ (44). In the present study, the Cronbach’s α of the OLBI was 0.85.
Statistical procedures
The analysis was conducted using SPSS version 25 for Windows (SPSS, Chicago, IL, USA) and R software (version 3.5.1, 2018-07-02). No replacement of missing values was performed since missing values were low and missing at random. Furthermore, missing values for the PCQ ranged from 0% to 5.6% and 0% to 0.6% for the OLBI. Cronbach’s alphas for the total scales were 0.685 for the SCQ, 0.872 for the PCQ 0.852, 0.757 for the burnout dimension exhaustion (EX) and 0.758 Disengagement (DIS).
T-tests and one-way between-groups analysis of variance were performed to investigate differences in SC and the two burnout dimensions depending on demographic variables. Chi-square and t-tests were used to examine associations and differences in dichotomised items from the PCQ concerning demographic variables. Effects were also analysed using the following effect size measures: Cohen's d (d), Eta-squared (η2), and Phi coefficient (φ). The criteria set by Cohen (1988) regarding effect sizes were used: small (d≥0.2, η2≥0.02, φ≥0.10), moderate (d≥0.50, η2≥0.06, φ≥0.30), and large effects (d≥0.80, η2≥0.14, φ≥0.50).
Partial least squares regression (PLSR) was used to identify items of importance in the SCQ and PCQ in relation to the two OLBI subscales of exhaustion and disengagement. The jackknife approximate t-tests of the regression coefficients were used to evaluate the independent importance of each item in the SCQ and PCQ to explain the variation in the response variables. Statistical significance was set at p <0.05. The jackknife 95% confidence intervals of the regression coefficients were used in the same manner to facilitate graphical interpretations of the findings. The number of components in the PLSR model was selected by examining the various validation plots.
Ethical considerations
This study was approved by the Organizational Joint Committee on Ethics in Biomedical Research, Tehran University of Medical Sciences (IR.TUMS.FNM.REC.2018-067). Participants were given verbal and written information about the study. They were assured that their participation was voluntary and that they could withdraw at any time without any explanation or consequences. All questionnaires were completed without noting the participants’ names or contact details, organizational characteristics, or any additional information that could violate their confidentiality. All the data were coded as they were entered into the datasheets. Printed copies of the questionnaires were saved in a locked cabinet at the first author’s workplace where no one else had access. All methods were carried out in accordance with relevant guidelines and regulations.