Dementia awareness among elderly at risk for developing mild cognitive impairment: a cross sectional study at a university-based primary care clinic

This cross-sectional study was conducted among elderly patients attending a university-based primary care clinic (Klinik Primer PPUKM Cheras, KPPC), which provides specialised primary care services to the surrounding communities within a 10-km radius, as well as supportive shared care initiatives with the main university hospital.

Elderly patients aged 60 years old and above under follow-up for chronic care, i.e., hypertension, diabetes mellitus, hyperlipidaemia, and gout management, were recruited. Elderly patients attending the outpatient clinic for other reasons and those with an established diagnosis of dementia or other cognitive impairment were excluded. The latter group was excluded on the assumption that their awareness of dementia would influence their responses in the questionnaire.

A simple random sampling method using a web-based application tool (stattrek.com) to select patients from the clinic’s Appointment Registry.

Potential patients (aged 60 years old and above) were identified from the daily attendance log of the Klinik Primer PPUKM Cheras (KPPC), a university-based primary care clinic in Kuala Lumpur. This formed the sampling frame for patient selection, as there was no clinic-based elder patient registry. Simple randomised sampling was conducted using a web-based application (Stattrek.com) to select patients from the sampling frame. Potential patients were contacted via a short messaging service using a poster informing them of the study. The poster included the official telephone number of the research team to avoid rejection of an unrecognised telephone number. The patients were contacted by the researcher after 24–48 h to assess their eligibility to participate. Patients who met the inclusion criteria were invited to participate in the study. Participation is voluntary, and refusal to participate or withdrawal from the study would not affect the treatment received from the clinic. Two scheduled reminders were given via text messaging over 1 week. Participants who did not respond to both reminders were considered as declined participation. Participants who agreed to participate were required to respond to the short messaging service. Upon receiving this notification, the researcher would then proceed to contact the participants via telephone. All prospective recruits were given a patient information sheet regarding the study and their role. A written informed consent (IC) was obtained from all subjects and/or their legal guardian(s) where applicable. The IC was included together with the self-administered web-based questionnaire which was distributed via mobile devices using the Google Form™ application. Refer Fig. 1 for details.

The sample size was estimated using the Leslie Kish formula, and a minimum sample of 227 participants was required, based on a precision of 0.05 and a prevalence of 16% from a study by Shahar et al. [20]. This total included an additional 10% dropout rate, which was added on to anticipate the withdrawal of participation.

The web-based self-administered questionnaire (Google Form™) consisted of four main sections; (i) socio-clinical demographics, (ii) awareness of dementia assessment, (iii) TUA-WELLNESS MCI risk screening tool, and (iv) consultation with healthcare providers. The questionnaire was bilingual (Malay and English) to accommodate participants with difficulty in either language. The time required to complete the questionnaire ranged from 5–10 min. In the case where assistance was required, the researcher will only assist by reading the questions verbatim through the phone.

In this study, awareness of dementia is defined as the ability to correctly answer basic knowledge statements on dementia, including a critical question, “Dementia is a part of the normal process of aging”, which the participants need to answer as false. The questions assessing dementia awareness were adapted from the Northern Ireland Life and Times (NILT) Survey 2010 [8]. The NILT 2010 is an annual survey which measures public opinion on identified themes and measures the changing public opinions since its inception in 1998 in Northern Ireland. The module on dementia was used in the 2010 survey (dementia knowledge and attitude).

Back-to-back translation of the questionnaire by Parkland et al. (2012) to the Malay language was conducted, and the contents and understandability of the questions were checked by Family Medicine Specialists who were well-versed and trained in community geriatrics and dementia care [8]. Only seven (7) out of 8 questions in Section 2 of the original NILT version covering dementia awareness were selected. During the validation process, the expert panel decided that one question on ‘drug treatments that help with dementia’ should be removed as the translated Malay version provided a similar meaning to ‘dementia can be cured’. The questionnaires consist of seven right or false statements that the participants need to answer.

The awareness of dementia was categorised into 3 levels based on the quartiles. A total score of 0 – 4, or inability to answer critical questions correctly, was categorised as low awareness; a total score of 5 indicated medium awareness, while a total score of ≥ 6 was categorised as high awareness. The total score of this section was used to look at factors associated with the awareness.

The risk of developing MCI in section (iii) of the questionnaire was assessed using the TUA (Towards Useful Aging) -WELLNESS tool developed locally and validated to address the multi-ethnic sociocultural differences in the local community [21]. It consists of 10 questions corresponding to items which cover education, fasting blood sugar (FBS) value, fasting cholesterol level, daily fruit/vegetable consumption, frequency of modern device usage, frequency of doing mechanical tasks for men, frequency of reading/sewing activity for women, frequency of practising calorie restriction, the difficulty faced in conducting daily activities, and their satisfaction of their quality of life. Each question contributes a score of 1 point, with a total score of 17. The risk of developing MCI is calculated from the total score. A score of lower than 11 indicates a high risk of developing MCI, and a score of 11 and more indicates a low risk.

A total of 268 elderly patients who met the inclusion and exclusion criteria were invited to participate. A total of 207 agreed and consented to participate, making the response rate 77.24% (207/268). Among the reasons for refusing to participate were: unable to complete the questionnaire (i.e. time constraints), language barrier and unwillingness to share personal medical information despite reassurance on confidentiality.

Table 1 displays the clinical and sociodemographic characteristics of study participants. The median age of participants was 70 (IQR 9) years.

Based on the total score of the TUA-WELLNESS questionnaire, 39.1% had a high risk of developing MCI, while the remaining 60.1% had a low risk.

The majority (92.8%) of the participants had a low level of dementia awareness (refer to Table 2).

On individual items (Fig. 2), only 19.8% of participants were able to correctly answer the critical statement (dementia is a part of the normal process of aging). Among 207 participants, slightly more than half (59.9%) incorrectly thought that dementia is a disease of the brain and almost half (49.8%) correctly stated that dementia is not a mental illness (refer Fig. 2).

Only 4.8% of the patients had some discussion about the risk of cognitive impairment with their physicians, while 8.2% were unsure. The majority of the participants (93.2%) did not initiate or share their concerns regarding dementia with their physicians, with 2.4% unsure if they had ever discussed this topic before.

However, gender, race, level of education, all chronic illnesses on their own and cardiovascular risk factor showed no significant differences in scores for awareness of dementia, as summarised in Table 3.

One of the limitations of this study is that potential participants were recruited via a telephone call. The participants had the option to not answer the telephone call at the point of recruitment. Opdenakker and colleagues cited that one possible reason could be the distraction by the environmental activities, rendering the interviewer at a disadvantage in creating a pleasant and conducive interview environment [42]. Other possible reasons would be refusing to talk to strangers, language barriers and long interview sessions. To overcome this issue, we sent an e-poster to each participant 2 days before conducting the telephone interview. This was done to introduce them to the researchers and the study purpose and allow the participants to prepare themselves for the interview session. Then, every time before starting the interview session, we informed the respondents that the interview would only require around 5 to 10 min to reduce the possibility of them rejecting or hanging up the phone.

Another limitation is that all clinical variables were self-reported by the respondents and not objectively measured. This may contribute to self-report response bias. Response bias, which stems from the respondent, has an impact on measurement quality by bringing error or a lack of precision into the study [43]. For various reasons, respondents may be unwilling or unable to provide accurate responses [43]. Future studies should include verification of medical conditions by accessing medical records.

Last but not least, participants’ responses may have been influenced by searching for information and answers regarding dementia on the internet since we had sent them a poster of our study prior to contacting them, and may contribute to bias.

Based on this study, people with chronic diseases at higher risk for MCI had limited knowledge of dementia. Efforts to increase awareness of the role of lifestyle factors in dementia risk reduction need to be urgently addressed, focusing on educating the public about the role of modifiable risk and preventive factors. Primary care physicians should address the elevated risk of MCI and subsequent dementia with their patients receiving chronic care to be better aware of the condition and leverage this information to increase their motivation to reduce their risk of developing dementia.

Future studies in multicentre trials should be conducted with patients and primary care doctors representing urban and rural communities to assess awareness levels and perhaps recommend targeted suggestions for improving awareness in these populations.