This analysis is based on data of a telephone-based questionnaire and semi-structured interview assessment in a convenience sample of n=141 people with cognitive impairment in the primary care setting in Mecklenburg-Western Pomerania and North-Rhine-Westphalia. Participants were recruited from currently running interventional trials intersec-CM (Supporting elderly people with cognitive impairment during and after hospital stays with intersectoral care management) [30] and the DCM:IMPact study (Dementia Care Management: Implementation into different Care Settings), an implementation study of the dementia care management of the DelpHi-MV trial [31-33]. Ethical approval has been obtained from the ethics committee of Greifswald Medical School (registry number: BB 159/17) and the ethics committee of the Chamber of Physicians of Westfalen-Lippe (registry number: 2017–688-b-S). Ethical approval for the DCM:Impact study was obtained by the ethics committee of the University Medicine Greifswald (BB 01/2019). Participants gave written informed consent to participate in these studies. In both studies, participants agreed to be regularly asked about their current living situation, health, quality of life, social contacts, health issues, burden, utilization of medical services and others. As such, conducting the assessments analyzed here were covered by the original ethical approval and no further approval was obtained. Written informed consent was obtained personally or by legal guardians when necessary. The assessments of the present analysis were conducted by trained study staff during the time of the lockdown during study contacts between 22nd of March and 5th of June 2020, before restrictions were lifted.
Sample
A total of n=115 participants of the intersec-CM trial received the additional assessment. Intersec-CM is a complex, longitudinal, multisite randomized controlled trial. It was designed to treat a hospital-based cohort of people above the age of 70 with an adaption of “Dementia Care Management” (DCM) – a treatment proven to be effective in primary care – to the discharge setting. As part of this, specifically trained study staff develops, implements, and monitors a treatment and care plan, based on comprehensive assessments during the hospital stay, recommendations at discharge and unmet needs at home. For the 3 months after discharge study staff coordinates treatment and care in close cooperation with the discharging hospital, treating physician and other care providers. The first patient from a total of n=401 was recruited in November 2018, the last in April 2020. Participants of the ongoing intersec-CM trial were asked to take part in the additional assessment only if they belonged to the original study’s intervention group or were part of the control group after completion of their individual follow-up time period.
A total of n=26 participants were interviewed in the framework of the DCM:IMPact study, a mixed-methods, multi-center, implementation study. DCM:IMPact compliments the DelpHi-trial (Dementia: Life- and person-centered help). In this study, the effective [32] and cost-effective [33] dementia care management is implemented in different care settings, including ambulatory care services, dementia care networks, and hospitals, aiming to reveal in which care setting the highest need and lowest implementation barriers for such model of care exist, and thus the highest effects could be achieved. The DCM:IMPact study started in 2019.
Owing to this particular ascertainment scheme (the localities in which the above studies were performed), participants come from two German federal states: Nordrhein-Westfalen and Mecklenburg-Vorpommern.
Data assessment
The benefit of interviewing a known convenience sample is that baseline sociodemographic characteristics and certain clinical variables are available. Variables that usually cannot be assessed by phone, like cognitive status, are known. The sample is described in detail in table 1. Additionally, study participants have an established relationship with the interviewers. We expect this personal contact to increase the validity of the data obtained. In case of uncertainty about the validity of the data provided by the person with cognitive impairment, the caregiver was asked to provide information..
From end of March until 6th of June, we conducted a telephone-based questionnaire and semi-structured interview assessment. The questionnaire included variables from previously validated instruments and is available as supplement (supplement 1). They were chosen based on their feasibility in telephone-based surveys, duration and feasibility in people with cognitive impairments. All were available in German versions and conducted in German. The decision for inclusion was based on an expert panel of researchers at the DZNE, the university of Greifswald, the university of Rostock, the university of Bochum and experts from the geriatric division of the hospital in Bielefeld. They cover:
- Attitudes, knowledge about and perceived personal impact of the pandemic
- social support
- loneliness
- anxiety
- depression
- change in the frequency of social activities due to the pandemic
- perceived impact of the pandemic on health care related services
Ad a) Participants were asked whether they were sufficiently informed about Corona (yes/ no) and whether they knew someone infected (yes/ no) and whether they rated the measures against the pandemic as appropriate ("Yes”, “No – too strict”, “No – not strict enough”). They were asked to rate whether they felt “1 – strongly”, “2 – moderately”, “3 – not at all” worried because of Corona for themselves; “stressed out”, “concerned about their health”, “the health of family and friends”, “worried about Corona in general”. They were asked to rate their fear to infect family members, whether their everyday life has changed by the pandemic and whether the visiting restrictions for nursing homes do affect their life. The items are listed in table 2.
Ad b) Social support was measured using selected items from the German version of the Resource Generator [34]. The Resource Generator covers a wide variety of social resources that can be divided in four aspects of social capital (capital associated to: prestige and education of the network, political and financial knowledge of the network, special skills of the network, social support of the network). For our analysis and due to time restrictions in the assessment four relevant items from the aspect “instrumental social support” were chosen (i.e. is there anyone, who “supports you when repairing something”, “supports you when shopping”, “supports you in filling out forms”, “supports you with legal and administrative challenges”). The participant was asked to rate these with: “0 – not at all”, “1 – rather not”, “2 – rather”, “3 – exactly”. Each item is analyzed separately and for the purpose of analysis, we add the scores of each item to generate a total score of social support, ranging from 0 to 12.
Ad c) Loneliness was measured using the loneliness scale of de Jong Gierveld & van Tilburg [35]. This scale consists of 6 items with statements the participant is asked to rate with “1 - exactly”, “2 - rather”, “3 - rather not” or “4 - not at all”. The scores of each item are summed to a total score, ranging from 6 to 24; the higher the score the higher loneliness is present in the participant. This instrument was chosen, because it is used in the longitudinal population-based cohort of the elderly in Germany and reference scores are available for the years 2008 to 2017 [36]. We furthermore measured loneliness with a focus on social relationship using a short loneliness scale developed specifically for use on a telephone survey [37]. Three items ask how often one feels “lacking companionship”, “being left out” and “being isolated”. Each item is rated “seldom”, “sometimes” or “often”. The sum score indicates the level of loneliness, ranging from 3 to 9.
Ad d) Anxiety is measured using the Generalized Anxiety Disorder Scale-7 (GAD-7) [38]. It is a screening instrument to diagnose a generalized anxiety disorder according to DSM-5 and ICD 10. The psychometric properties are considered appropriate. The German version is validated for the general population [39]. Participants are asked to rate how frequent symptoms of anxiety have occurred during the time-frame of 2 weeks preceding the interview (“0 – not at all”, “1 – single days”, “2 – more than half of the days”, “3 – nearly daily”). The GAD-7 delivers a sum score by adding each item and ranges from 0 to 21. The sum score yields whether anxiety symptoms are minimal (0-4), low (5-9), medium (10-14) or severe (15-21). Furthermore, there are cut-off scores for a generalized anxiety disorder being possible (score >10) or probable (score>15).
Ad e) The two item Patient Health Questionnaire (PHQ-2) was used for screening for depressive symptoms [40, 41]. The participant is asked to categorize how often he/ she has experienced “loss of interest or pleasure in activities” and “sadness” during the 2 weeks prior to the assessment. The options are “0 – never”, “1 – single days”, “2 – more than half of the days”, 3 – nearly daily”. These two items represent the main indicators for a positive screening of depression according to DSM-5, with one of both being mandatory. Depression was considered possible if one of the two items was scored >= 2. Depression was considered unlikely otherwise. For the sample of participants recruited for the intersec-CM study we included pre-pandemic data and could therefore calculate whether there was a change in frequency of symptoms during the period of lockdown (“more frequent”, “equally frequent”, “less frequent”).
Ad f) To measure change in frequency of (social) activities due to the pandemic we use a generic list of social activities and ask whether the specific activity is carried out “more frequently”, “less frequently” or “with no change in frequency” during the time of lockdown. If an item does not apply to the participant´s situation, it was marked as being “not applicable” (e.g. if a participant never “went dancing” previously, this item was marked as “not-applicable”). The list consists of 24 items (see table 3). Each item is analyzed separately. We furthermore provide three scores per person, which represent the number of activities that have “increased”, “decreased”, and been “unchanged” in frequency.
Ad g) To assess the perceived impact of the pandemic on health care related services participants are asked to rate whether the provision and use of the service is “better”, “worse” or “unchanged” during the lockdown in comparison to the time before. Items that were irrelevant to a participant’s situation were marked “not applicable” (e.g. if a participant never utilized “podiatry” as a service, this item was marked as “not applicable”). The full list is provided in table 4. Each service is analyzed separately. We furthermore provide three scores per person, which represent the number of services where the quality is rated to be “better”, “worse” and “unchanged”.
Statistical analysis
In this exploratory study we provide descriptive statistics for the variables under analysis. Descriptive analyses were conducted using IBM SPSS Statistics 21. Participants with missing data in single items were excluded from the analysis of this special item or the scale this item is part of. No data were imputed.