The relationship between disability and depression is complex, and previous studies showed that the relationship may be mediated by positive social interaction. The main objective of this study was to examine whether positive social interaction mediates the association between disability and depression in older adults. The data of this analysis were drawn from 2322 community-dwelling older adults aged 60 to 92 years in Peninsular Malaysia who participating in the baseline Neuroprotective Model for Health Longevity (TUA) study. Depression was measured by Geriatric Depression Scale, disability was assessed by World Health Organization Disability Assessment Schedule 2.0, and positive social interaction was measured by 4 items under the positive social interaction domain of The Medical Outcomes Study Social Support Survey. Hierarchical Multiple Linear Regression was performed by using SPSS version 23.0 to examine the mediation effect of positive social interaction. Next, Sobel Test was used to validate the mediation effect. Results showed that both disability (β = 0.086, p < 0.001) and positive social interaction (β = -0.107, p < 0.001) significantly predicted depression in the final model, after controlling for possible confounders (gender, marital status, year of education). Of most interest, positive social interaction was found partially mediated the association between disability and depression (from β = 0.094, p < 0.001 to β = 0.086, p < 0.001). Furthermore, significant Sobel Test (z = 2.519, p = 0.012) confirmed the mediation effect of positive social interaction. These findings reinforce the role of disability and positive social interaction in predicting mental health in old age. To prevent depression in old age, specific intervention to maximize the positive social interaction among disabled older adults is warranted.
The increased use of health care services by elderly has placed greater pressure to an already strained health care resources. Thus, an accurate economic cost estimation for specific age-related diseases like dementia is essential. The objectives of this project are to estimate costs of treating patient dementia among Malaysian elderly in the hospital settings. Two types of data were collected: Hospital costing data (using costing template) and patient clinical data (using questionaire). The cost analysis for hospital setting was carried out using a step-down costing methodology. The costing template was used to organize costing data into three levels of cost centers in hospitals: overhead cost centers (e.g. administration, consumables, maintenance), intermediate cost centers (e.g. pharmacy, radiology), and final cost centers (all wards and clinics). In estimating the cost for each cost center, both capital cost (building, equipment and furniture cost) and recurrent cost (staff salary and recurrent cost except salary) were combined. Information on activities which reflects the workload such as discharges, inpatient days, number of visit, floor space etc., are gathered to determine an appropriate allocation factor. In addition, for each final cost center, the fully allocated costs are then divided by the total unit of in-patient days to obtain the cost of providing services on a per-patient per-day of stay basis, referred as unit cost. The unit cost is finally multiplied with the individual patient’s length of stay to obtain the cost of care per patient per admission. All these steps were simplified by using the Clinical Cost Modeling Software Version 3.0 (CCM Ver. 3.0). The mean cost of dementia cases per episode of care was RM 12,806 (SD=10,389) with the length of stay of 14.3 (SD=9.9) days per admission. The top three components of cost for the treatment of dementia were the ward services 8,040 (SD=7,512), 62.78% of the total cost, followed by the pharmacy 1,312(SD=1,098), 10.25% of the total cost and Intensive Care Unit 979 (SD=961), 7.64% of the total cost. A multivariable analysis using multiple linear regressions showed that factors which significantly influence (p<0.05) the treatment costs of dementia cases were the length of stay (p<0.001), followed by age (p=0.001), case type severe (p=0.005) and study location (p=0.032). However, the factor length of stay is the tremendous parameter. In conclusion, data collection from selected hospitals as well as patient level data from medical record unit were successfully used to estimate the provider costs of hospital for the elderly with dementia disease. Results from the project will enable an assessment on the economic impact and consequences of cognitive impairment in an aged population. A cost quantification and distributive mapping of the burden of care can assist in policy implementation through targeted intervention for at-risk groups, which will translate into savings by means of delayed onset or progression of dementia.