Older adults' health is one of the most concerning world issues. Their healthcare has gained more importance in recent decades because of medical advances and the increase in the older population (Ghasemi et al, 2017). In Iran, the proportion of the population aged over 60 years is rising continually, and accounted for 9.35% of people in the country in 2016 (about 7.5 million people) (Statistical Center of Iran (SCI), 2018).
A general concern around the care of older adults is patient safety (Williams and Wold, 2015). Factors such as frailty, disease and polypharmacy (Palmer and Roessler, 2000) as well as complexity of care raise the risk of undesirable events in older adults (Le Pogam et al, 2015). Evaluation of safety hazards and making environmental modifications can improve safety (Williams and Wold, 2015). Self-advocacy has been suggested as one of the most effective approaches in promoting older adults' safety (Bishop, 2012; Elder et al, 2008; Jonikas et al, 2013).
Self-advocacy means participating in health care and becoming active in the treatment process. It includes knowing your strengths and needs, knowing your rights, setting goals and being able to communicate with others about them (Tuttle and Silva, 2007). Patient participation in designing self-care is considered a self-advocacy outcome. Self-advocacy means collecting and applying information for health promotion (Elder et al, 2010; Hagan and Donovan, 2013a). Skills required for self-advocacy include communication, information-seeking, problem-solving, decision-making, negotiation and bargaining (Wiltshire et al, 2006; Vaartio, 2008).
Self-advocacy is a key strategy of self-management (Vaartio, 2008), and self-management is an outcome of self-advocacy (Hagan et al, 2017).
Factors that affect self-advocacy capacity include knowledge, cognitive abilities, personality, developmental level, age, motivation, social variables (such as economic and social roles and situation) and healthcare organisation variables (such as philosophy of care) (Naumann and Vessey, 2002), as well as variables such as disease, low literacy, cultural habits and socialisation (Elder et al, 2008).
Although older adult patients are at a greater risk of medical errors, they may not have the ability to advocate for themselves or feel comfortable taking on the self-advocacy role (Elder et al, 2008). Older adults are more often exposed to medical errors and safety hazards because they use healthcare services more than others, are less likely to participate in medical decision-making (Levinson et al, 2005; Belcher et al, 2006) and have low rates of treatment adherence (Bosworth, 2010).
Empowering older adults to be self-advocates in patient safety can lead to more hope and participation in their health care (Pickett et al, 2012) and, consequently, can influence clinical outcomes and patient safety (Elder et al, 2007). It can facilitate them becoming more and better engaged in decision-making and taking control of their treatment (Pickett et al, 2012).
Elder et al (2008) showed that self-advocacy in older adults can be promoted by providing a training intervention about medical errors, teaching communication skills so they can interact better with medical practitioners, and encouraging them to practise patient safety recommendations.
Patient safety is a nursing outcome and its maintenance is a result of nursing care (Liu et al, 2014). As patient advocates, nurses play an important role in ensuring the safety and quality of care (Ronnebaum and Schmer, 2015; Cho et al, 2016) and effective nursing advocacy can potentially increase patient safety (Hanks, 2010). A culture of safety motivates nurses to mitigate risks and hazards in patient care (Groves and Bunch, 2018) and nurses have special responsibility to patients with low levels of self-advocacy (Negarandeh et al, 2006).
Reviewing literature from recent years shows that there is little evidence about older adults' participation in patient safety and their embracement of patient safety self-advocacy behaviours, especially in developing countries. PubMed, Scopus, Embase, ProQuest, SAGE journals, Springer, MEDLINE, Elsevier (Science Direct), Ovid, Google Scholar and Web of Science were searched using the keywords ‘patient safety’, ‘elderly’, ‘seniors’, ‘old patient’, ‘elder’, ‘self-advocacy’, ‘patient participation’, ‘patient involvement’ and ‘patient engagement’ in varying combinations. Abstracts of research studies, reviews and conferences, editorials and books published in English between 2008 and 2018 were reviewed. This search resulted in 171 pieces of literature. The abstract review identified nine relevant, original articles that were focused on older adult patients' participation in patient safety or their embracement of patient safety self-advocacy behaviours. All results were restricted to developed countries; the authors searched for, but could not find, any research based in a developing country.
This study aimed to measure older adults' participation in and embracement of patient safety self-advocacy behaviours, and tested the research hypothesis that these and their components correlated with some demographic characteristics.
Material and methods
Participants
This cross-sectional descriptive study was conducted in 2017. The participants were 230 patients aged 60 years and above who were hospitalised in government and social security hospitals in Hamadan, Iran (n=5).
As the hospitals had different specialties and the patients had different socioeconomic status, participants were selected using a stratified random sampling method to make them as representative as possible of the general 60+ population (Polit and Beck, 2004). The total sample size was calculated based on the monthly admission rate of elderly patients (aged 60 years and above) in all hospitals and the sample size was divided proportionally for each hospital. Variance was estimated based on a pilot study. Random selection of elderly patients was done based on the daily admission list of hospitals in the study period.
Inclusion criteria were having been hospitalised at least once within the last year and having had a minimum of three visits to a doctor within the previous 6 months. Exclusion criteria were cognitive disease and the inability to answer the questions. Given that only 46.5% of older people in Iran are educated (SCI, 2018), illiteracy was not an exclusion criteria. The questions were read out to the illiterate participants by a trained researcher.
Finally, 268 hospitalised patients were selected; 32 were excluded because they decided not to take part, had a poor medical condition, cognitive disease or a language barrier. Six questionnaires were not completed or were not returned (97.2% response rate). This left 230 participants.
Data
The data collection instrument used for this research was the Persian version of a survey instrument for measuring older adults' participation and embracement of self-advocacy in patient safety behaviours, called the Seniors Empowerment and Advocacy in Patient Safety (SEAPS) survey (Elder et al, 2007). This instrument is a self-report questionnaire that consists of 21 questions, which measure older adults' participation and embracement of self-advocacy in patient safety behaviours on four subscales based on the health belief model. Table 1 shows the subscales and how the scoring works. The psychometric testing of this instrument was done in the Midwest area of the US (Elder et al, 2007).
| Subscale | Description | Meaning | Items | Item score | Number of items | Total score |
|---|---|---|---|---|---|---|
| Behaviour | A self-report proxy for the performance of actions | How often the patient carries out behaviours related to recommended safety tasks | Safety tasks recommended by Elder et al (2007) included: teaching themselves about health problems and medicines; asking doctors about health problems, laboratory tests and medicines; keeping an updated list of all medicines; checking on results of laboratory or X-ray tests; getting a second opinion from another doctor when needed; asking a friend or family member to accompany the patient during doctors' visits; feeling comfortable to change the physician if health concerns were not addressed; providing the doctor with a complete history of the disease; complaining in writing if the patient had a problem with care providers (these last three safety recommendations were not considered in the behaviour subscale and were items in the attitude, self-efficacy and outcome-efficacy subscales respectively) | 4–point Likert scale from 0–3 | 6 | 0–18 |
| Self-efficacy | Confidence in own ability to take action effectively | How confident patients are that they could carry out the recommended safety tasks | 4–point Likert scale from 0–3 | 5 | 0–15 | |
| Attitude | Concerns about barriers to participating in the actions | How much patients agree or disagree with the recommended safety tasks | 4–point Likert scale from 0–3 | 5 | 0–15 | |
| Outcome efficacy | The belief that the actions will benefit health | Belief that doing the recommended safety tasks will improve their health | 4–point Likert scale from 0–3 | 5 | 0–15 | |
| Total | 21 | 0–63 |
Source: Elder et al, 2007; Novinmehr, 2019
The Persian version of this instrument was developed after translation and cultural adaptation and psychometric study in Tehran, Iran (Novinmehr, 2019). Content validity of the translated survey was measured by an expert panel. Face validity was measured by a sample of older adult patients. A total of 16 men (80%) and 4 women (20%) participated (n=20); 70% were living in cities. The mean age was 69.9 years, and participants had had a mean 7.0 years of education. The impact score of any question was measured by an impact method described by Juniper et al (1997) and Lacasse et al (2002). Each participant assigned a score of between 1 and 5 to each question about the appropriateness of the item; the authors then calculated the mean scores and the frequency of scores 4 and 5. The authors then multiplied the mean by the percentage of frequency and thus calculated the impact score for the 21 questions. Reliability was confirmed by Cronbach's alpha (α=0.887) and the Wilcoxon test (P=0.666) (Novinmehr, 2019).
The self-advocacy score and its subscales were divided into ranges of very low, low, moderate, high and very high by intervals of 12.6 for self-advocacy, 3.6 for behaviour, and 3.0 for self-efficacy, outcome efficacy and attitude (Novinmehr, 2019).
Data analysis
The assumption of normal distribution of all variables was tested using the Kolmogorov-Smirnov test. Twelve independent variables—age, years of education, number of children, number of cohabitants, frequency of visits, number of hospital admissions in the previous year, lifetime total number of hospitalisations, sex, marital status, rural living, cohabitants and occupation—were considered demographic characteristics. To identify any association between the dependent variables (self-advocacy and its subscales) and demographic characteristics, an analysis was performed using Pearson correlations, an independent t-test, ANOVA testing and post-hoc testing using the Tukey method. To study the statistically significant effects of demographic characteristics on self-advocacy and its subscales, multiple linear regressions were undertaken. Data analysis was conducted using SPSSv16 and Excel 2003.
Ethical approval
This study was approved by the ethical committee of the deputy of research and technology of Tehran University of Medical Sciences (code: IR.TUMS.REC.1394.621). The researcher and research goals were introduced to the participants; those who gave informed consent were then provided with questionnaires.
Results
Sample
Participants' demographic characteristics are shown in Table 2. A total of 119 men (51.7%) and 111 women (48.3%) participated (n=230); the mean age was 69.9 (SD=8.2; range 60-96 years), 70.9% were married, 59.1% were living in cities and 61.3% kept house or were retired. The majority (58.7%) were aged over 65 years old; 78.5% had more than three children, 82.9% were not living alone and 60% were illiterate.
| Frequency | Percentage | |
|---|---|---|
| Sex | ||
| Male | 119 | 51.7% |
| Female | 111 | 48.3% |
| Occupation | ||
| Housewife | 96 | 41.7% |
| Retired | 45 | 19.6% |
| Private | 30 | 13.0% |
| Disabled/jobless/etc | 59 | 25.7% |
| Marriage status | ||
| Married | 163 | 70.9% |
| Widowed/single | 67 | 29.1% |
| Living area | ||
| Urban | 136 | 59.1% |
| Rural | 94 | 40.9% |
| Age (years) | ||
| 60–65.9 | 95 | 41.3% |
| 66–71.9 | 43 | 18.7% |
| 72–77.9 | 44 | 19.1% |
| 78–83.9 | 30 | 13.0% |
| 84–89.9 | 15 | 6.5% |
| 90–95.9 | 2 | 0.9% |
| ≥96 | 1 | 0.4% |
| Years of education | ||
| 0 | 138 | 60.0% |
| 1–5 | 41 | 17.8% |
| 6–8 | 21 | 9.1% |
| 9–12 | 19 | 8.3% |
| >12 | 10 | 4.3% |
| Living in same house (cohabitant) | ||
| Spouse | 152 | 66.1% |
| Children | 40 | 17.4% |
| Alone | 37 | 16.1% |
| No answer | 1 | 0.4% |
| Living in same house (number) | ||
| 0 | 39 | 17.0% |
| 1–2 | 115 | 50.0% |
| 3–4 | 53 | 23.0% |
| 5–6 | 16 | 7.0% |
| 7–8 | 5 | 2.2% |
| No answer | 2 | 0.9% |
| Number of children | ||
| 0 | 8 | 3.5% |
| 1–3 | 41 | 17.8% |
| 4–6 | 98 | 42.6% |
| 7–9 | 70 | 30.4% |
| 10–12 | 11 | 4.8% |
| No answer | 2 | 0.9% |
| Number of doctor visits in 6 months | ||
| 3–9 | 179 | 77.8% |
| 10–16 | 40 | 17.4% |
| 17–23 | 6 | 2.6% |
| 24–30 | 4 | 1.7% |
| 31–37 | 0 | 0.0% |
| ≥38 | 1 | 0.4% |
| Hospitalisations in last year | ||
| 1–2 | 184 | 80.0% |
| 3–4 | 33 | 14.3% |
| 5–6 | 9 | 3.9% |
| 7–8 | 2 | 0.9% |
| 9–10 | 2 | 0.9% |
| Lifelong hospitalisations | ||
| 1–6 | 169 | 73.5% |
| 7–12 | 40 | 17.4% |
| 13–18 | 6 | 2.6% |
| 19–24 | 12 | 5.2% |
| 25–30 | 1 | 0.4% |
| ≥31 | 2 | 0.9% |
Measurement of self-advocacy
The self-advocacy score average was 40.16 out of a maximum of 63 (SD=9.6; range 9-57). Two (0.9%) respondents had a very low average score; 16 (7%) had a low average score; 75 (32.6%) had a moderate score; 118 (51.3%) had a high average score; and 19 (8.3%) respondents had a very high score (Figure 1).
The behaviour score average was calculated as 10.5, which was within the moderate range, from a total score of 18 (SD=0.2; range 2–18). The self-efficacy score average was at 10.2 out of a total score of 15 (SD=0.2; range 0-15). The attitude score average was recorded as 9 out of a total of 15 (SD=0.2; range 1-15). Outcome efficacy score average was at 10.5 out of 15 (SD=0.2; range 1-15).
A minority of respondents (41.35%) reported that they embraced self-advocacy in patient safety behaviour. Only 25 respondents (10.9%) noted that they always kept an updated list of all medicines, 47 (20.4%) said they always got a second opinion from another doctor when needed, 63 (27.4%) reported that they always taught themselves about health problems and medicines, 123 (53.5%) said that they always asked doctors about health problems, 149 (64.8%) reported they always asked a friend or family member to accompany them to a doctor visit and 163 (70.9%) said they always checked the results of laboratory or x-ray tests.
Relationship with demographics
Table 3 presents the relationships between demographic variables and scores for self-advocacy and its subscales. Self-advocacy scores had an inverse relationship with age (P=0.0 09) and a direct relationship with years of education (P>0.001). Self-advocacy scores had a significant relationship with sex (P=0.037), rural living (P=0.001) and marital status (P=0.004).
| Variable | Test | Behaviour | Self–efficacy | Attitude | Outcome efficacy | Self-advocacy score |
|---|---|---|---|---|---|---|
| Age | Pearson correlation (r) | –0.158 (0.017) | –0.148 (0.025) | –0.141 (0.033) | –0.089 (0.177) | –0.172 (0.009) |
| Years of education | Pearson correlation | 0.224 (0.001) | 0.228 (0.001) | 0.229 (<0.001) | 0.178 (0.007) | 0.273 (<0.001) |
| Number of children | Pearson correlation | –0.078 (0.239) | –0.065 (0.325) | –0.056 (0.401) | –0.093 (0.160) | –0.091 (0.170) |
| Living in same home | Pearson correlation | –0.019 (0.777) | 0.064 (0.339) | –0.015 (0.816) | 0.066 (0.319) | 0.027 (0.690) |
| Frequency of visits | Pearson correlation | –0.021 (0.755) | –0.012 (0.862) | –0.082 (0.218) | –0.035 (0.599) | –0.046 (0.487) |
| Last year's hospitalisation | Pearson correlation | 0.067 (0.313) | –0.032 (0.631) | –0.073 (0.271) | –0.038 (0.564) | –0.020 (0.768) |
| Lifelong hospitalisation | Pearson correlation | 0.064 (0.332) | 0.011 (0.868) | 0.000 (0.995) | 0.017 (0.803) | 0.031 (0.643) |
| Sex | t* (DF=228)** | 0.851 (0.396) | 1.504 (0.134) | 2.662 (0.008) | 1.765 (0.079) | 2.097 (0.037) |
| Marital status | t (DF=228) | 1.658 (0.099) | 2.214 (0.028) | 2.809 (0.005) | 2.747 (0.006) | 2.937 (0.004) |
| Rural living | t (DF=228) | 2.362 (0.019) | 3.073 (0.002) | 2.633 (0.009) | 2.775 (0.006) | 3.430 (0.001) |
| Living in same home | F*** (DF=228) | 1.070 (0.345) | 2.785 (0.064) | 2.805 (0.063) | 4.425 (0.013) | 3.809 (0.024) |
| Occupation | F (DF=229) | 4.639 (0.004) | 2.082 (0.103) | 3.092 (0.028) | 2.168 (0.093) | 4.468 (0.005) |
One-way ANOVA revealed a relationship between self-advocacy score and occupation (P=0.005). To establish a more accurate interpretation of variance analysis, the post-hoc test was carried out using the Tukey method. Results showed that self-advocacy scores in the retired group were significantly higher than other occupational groups, such as housewives (P=0.03), those in private occupation (P=0.042) and those who were disabled/jobless (P=0.03).
According to the results of the one-way ANOVA test, the self-advocacy score is associated with older adults' cohabitants (living with a spouse, children, alone etc) (P=0.024). The post-hoc test was carried out using the Tukey method to improve accuracy of the analysis of variance. Results showed that older adults who lived with their spouse had higher self-advocacy scores than those who lived alone (P=0.03). The post-hoc test showed that the self-advocacy score in older adults who lived with their children was not significantly higher than in those who lived alone (P=0.666) and was not significantly lower than in those who lived with their spouse (P=0.227). Based on the results, there was no significant relationship between self-advocacy score and number of children, number of cohabitants, the frequency of visits, hospitalisation in the previous year and lifelong hospitalisation rate.
Multiple linear regressions
Multiple linear regression tests were performed on all variables that had a significant relationship with scores of self-advocacy and its subscales in an independent t-test, Pearson or ANOVA. Age, years of education, cohabitants, rural living, occupation, marital status and sex were entered in this model as predictor variables and self-advocacy score as the outcome variable. Multiple linear regression tests were repeated for each subscale of self-advocacy as the outcome variable, but some of the mentioned predictor variables were absent.
Rural living was a significant predictor of total self-advocacy score (β=-0.168; P=0.016), age (β=-0.146; P=0.031) and occupation (retirement) (β=0.195; P=0.025) were significant predictors of the behaviour subscale score. Rural living was a significant predictor of the self-efficacy subscale (β=-0.170, P=0.015), sex was a significant predictor of the attitude subscale (β=-0.258; P=0.036) and rural living was a significant predictor of outcome efficacy subscale (β=-0.174, P=0.013).
Table 4 shows the results of multiple linear regression tests on the association of demographic variables with self-advocacy score and its subscales.
| Predictors | Standardised β | P value | Standardised β | P value | Standardised β | p–value | Standardised β | P value | Standardised β | P value |
|---|---|---|---|---|---|---|---|---|---|---|
| Age of respondent | –0.132 | 0.057 | –0.146 | 0.031 | –0.095 | 0.158 | –0.117 | 0.093 | *** | *** |
| Respondent's sex | –0.120 | 0.326 | *** | *** | *** | *** | –0.258 | 0.036 | *** | *** |
| Years of education | 0.117 | 0.142 | 0.073 | 0.370 | 0.127 | 0.077 | 0.117 | 0.144 | 0.086 | 0.223 |
| Marital status | –0.137 | 0.357 | *** | *** | –0116 | 0.086 | –0.114 | 0.123 | –0.089 | 0.537 |
| Rural living | –0.168 | 0.016 | –0.084 | 0.233 | –0.170 | 0.015 | –0.136 | 0.054 | –0.174 | 0.013 |
| Spouse as cohabitant | 0.045 | 0.760 | *** | *** | *** | *** | *** | *** | 0.166 | 0.264 |
| Children as cohabitant | 0.076 | 0.363 | *** | *** | *** | *** | *** | *** | 0.096 | 0.256 |
| Housewife occupation | 0.088 | 0.478 | 0.033 | 0.679 | *** | *** | 0.132 | 0.292 | *** | *** |
| Retired occupation | 0.088 | 0.302 | 0.195 | 0.025 | *** | *** | –0.001 | 0.989 | *** | *** |
| Private occupation | –0.070 | 0.346 | 0.029 | 0.700 | *** | *** | –0.145 | 0.052 | *** | *** |
Multiple linear regression tests were performed on all variables possessing a meaningful relationship with self–advocacy score and its subscales in an independent t–test, Pearson or ANOVA. The variables of age, educational level, cohabitants, rural living, occupation, marital status and sex were entered in this model as predictors and self–advocacy score as the dependent variable. Some predictors that had no a meaningful relationship were not entered in the models of its subscales.
Discussion
Few studies have been conducted on self-advocacy regarding patient safety of older adults, especially in developing countries. Based on the data from 230 older adult patients, we found that the total score of self-advocacy in patient safety is high and there is a significant relationship between a low self-advocacy score and living in a rural area. The subscale scores are associated with age, occupation, rural living and sex.
Measuring self-advocacy showed the majority of participants believed that self-advocacy behaviours can benefit their health (outcome efficacy) and were confident about their ability to adapt those practices (self-efficacy), and comfortable they could overcome the barriers to self-advocacy (attitudes); however, the majority of them did not report that they always carry out the self-advocacy behaviours. It was observed that there were high and very high scores related to self-advocacy and its subscales in the majority of respondents except for the behaviour subscale score.
These findings are in line with previous reports, such as a study that measured self-advocacy scores in elderly American patients, using the original version of the SEAPS instrument and then developing a patient-centred intervention to train them to be self-advocates for safer care (Elder et al, 2008). The majority of respondents, who were elderly African Americans, expressed a high level of self-advocacy (Anthony, 2007).
No evidence was found in the literature reviewed about self-advocacy scores in developing countries. However, given the problem of low levels of literacy in some countries (Samadi and McConkey, 2011) or unfamiliarity with democratic styles of functioning (Thomas and Thomas, 2005) and inadequate patient involvement in patient safety (Siddiqi et al, 2012), high self-advocacy scores in elderly patients in a developing country would not be expected.
Although several studies suggest that elderly patients are passive during medical encounters and have low health literacy (Kahana et al, 2009; Wolf et al, 2005), other findings differ. Recent studies showed that cognitively capable older patients can play an active role in managing their health (Ruggiano and Edvardsson, 2013). Some variables, such as increasing media use by older adults as a trusted source of health information, may affect their knowledge, attitudes and behaviour (Tennant et al, 2015) and these changes are not limited to developed countries. Health literacy is a key determinant of health and relates to the use of health information in decision making (de Wit et al, 2018). Education has a direct correlation with health literacy in the elderly (Prabsangob et al, 2018; Smith et al, 2018), and this finding was supported by a study on Iranian elderly women (Mahdizadeh and Solhi, 2018). Nevertheless, it may be valuable to investigate the circumstances that may lead to adequate levels of health literacy in illiterate older adult patients despite accessible information not being readily available and difficulties in health-information seeking.
A central issue is how older adults can change their behaviours to become an active self-advocate in patient safety. An older patient's health beliefs may gradually change because they receive more information about their safety, but may face further challenges regarding behaviour change. The adoption of some self-advocacy behaviours to support patient safety (such as keeping an updated list of all medicines and getting a second opinion from another doctor when needed) had the lowest score. Keeping an updated list of all medicines needs literacy, which is a major challenge in developing countries, especially in those where the elderly population is illiterate.
In another study on personal medication lists, only 38% of patients reported using these (Chae et al, 2009). Older patients need supportive, alert family members to help them keep and update their list and present it during medical visits. The physician who cares for the patient must also remind the older person to show them their list. A lack of reinforcement by health professionals is a barrier to personal medication record use (Traynor, 2007).
Getting a second opinion from another doctor when needed is a self-advocacy behaviour that was not reported by the majority of respondents. This action requires adequate information and critical thinking skills. The oriental culture encourages appreciativeness and contentment. Therefore, the patient's trust in the doctor is a traditional value that lowers the tendency to get a second opinion in Iran. Another barrier may be the costs of additional visits. Although patients acknowledge the value of second opinions, only one in six actually asks for one (Payne et al, 2014).
The research hypothesis that self-advocacy and its subscales correlate with some demographic variables was tested. Since correlations and associations between self-advocacy score/subscales and demographic variables may be affected by some confounding variables, multiple linear regressions were used.
Results showed that, among all the variables that have a relationship with self-advocacy score (age, years of education, occupation, marital status, cohabitants and rural living), only rural living was found to be a significant predictor of the self-advocacy score; the findings suggest that living in a village may be negatively correlated with self-advocacy. A study that was not limited to older adults found there was no significant difference in self-advocacy scores between people who live in cities and those who live in villages (Elder et al, 2010). Ease of access to medical services and information may have an effect on urban patients' self-advocacy. It may therefore be concluded that older patients in rural areas may have more need for interventions to be a more successful self-advocate. Conducting a qualitative study may help to test this observation.
Similarly, rural living was found to be a significant predictor of self-efficacy. The self-efficacy score was higher in people who lived in cities than in those who lived in villages. Previous studies have shown rural patients have low scores of self-efficacy (Young et al, 2017). Another study, which was conducted in Iran, found low and moderate self-efficacy in rural patients (Kamran et al, 2014). The self-efficacy score reflects how confident an older adult is in taking action effectively as a self-advocate and adopting patient safety behaviours, and may predict how much they persist with such practice (Elder et al, 2008; Kamran et al, 2014). This observation may be explained by poor literacy and low socioeconomic status among rural patients.
One of the more important findings from this study is that age is a significant predictor of behaviour. A lower behaviour subscale score was found in older adults with more advanced age, but this negative correlation is weak. It is expected that an older patient would have more difficulties in getting involved in their care. This partly depends on their knowledge and is significantly related to their physical and cognitive capacities (Davis, 2004; Levinson et al, 2005; Belcher et al, 2006); obstacles to self-advocacy, such as hearing problems, communication disorders and social inactivity, are likely to increase with age. In a study on women with cancer, younger women had higher self-advocacy scores than older women (Hagan and Donovan, 2013b). In another study, patients with HIV were categorised into two (active and inactive) groups in terms of self-advocacy and social activism; there was a significant difference between the two groups in terms of age (Brashers et al, 2002). This finding is consistent with a study on ambulatory patients (Elder et al, 2008). This may indicate that older adults of more advanced age may experience more barriers to behaviour as self-advocates in patient safety. Surprisingly, the correlation between age and self-advocacy behaviour was weak in the present authors' study. A key point is that the authors assessed this correlation in adults aged between 60 and 96 years, of whom 80% were under 78; they did not assess this correlation in the general population and across all age groups. This explains the weakness of correlation.
Frequent doctor visits are expected to have an impact on the abilities and beliefs of an older adult in advocating for themselves, but the findings suggest that there is no observable relationship between the frequency of doctor visits and self-advocacy score. This supports the findings of a study of older American adult patients (Elder et al, 2008).
It is notable that the findings of the present study are somewhat in contrast to research that identified an observable significant difference in self-advocacy in medical visits among married and unmarried patients (Elder et al, 2010). The findings of the current study showed that being married was significantly associated with self-advocacy. However, in the presence of other variables, this is not considered a significant predictor of self-advocacy. It may therefore be concluded that its association with self-advocacy may be because of some confounding variables such as loneliness and support networks. Future research is needed to investigate this relationship more fully.
Limitations and strengths
A decisive cause and effect relationship between demographic variables and self-advocacy score or its subscales cannot be confirmed because this is a cross-sectional descriptive study. Because a self-reporting instrument was used, answers given may be affected by how acceptable people thought they would be, for example regarding making written complaint against medical personnel and changing doctors. Caution is therefore needed regarding the generalisation of the results.
The fact that there are few studies on older adults' self-advocacy in patient safety, the presence of sensory and cognitive limitations and the effect of variables such as character, intelligence, support networks and socioeconomic status on the embracement of self-advocacy behaviour are also limitations to this research. These limitations were reduced as much as possible by applying exclusion criteria and using stratified sampling. Integrating these results with qualitative studies can be helpful in data enrichment.
Using a sample that is representative of the population and of an adequate size gives this study an advantage in reliability over similar research conducted on this topic.
Conclusion
This study contributes to the understanding of self-advocacy and its findings and could be used to help design more effective approaches to empowering older adults to be self-advocates in patient safety.
This cross-sectional study measured self-advocacy in patient safety and its subscales, and their relationship with demographic variables in older adult patients in Iran, a developing country. Unexpectedly, it showed a high average self-advocacy score. However, some self-advocacy behaviours such as teaching themselves about health problems and medicines, keeping an updated list of all medicines and getting a second medical opinion when needed were not consistent with the safety recommendations, and should be promoted.
Older adults may need interventions to reinforce their self-advocacy in patient safety. The need for this may be higher in rural patients and older adults of more advanced age because they showed lower self-advocacy. Therefore, conducting research into such interventions appears to be relevant. Investigating nurses' role in training, facilitating and empowering self-advocacy behaviour in older adult patients and studying the effectiveness of this training in different societies would be helpful.