***What does this study of frail elderly have to do with children? one reader asks. Not much. But it's a good example of quasi-experimental study design, which will help with developing research methodologies for all populations.***
Integrated service delivery (ISD) is a coordinated model emphasizing social interventions based on cooperation between multiple levels of service planning for frail elderly. In their study, Tourigny et al. (2004) evaluated the effectiveness of an ISD network of health and social services for frail elderly living in two semi-urban communities in Quebec. The sample included people 75 years and older (n=482) from two communities. Individuals from Bois-Francs served as the experimental group (n=272) and participants from Drummondville, where there were no ISD services available, served as the control group (n=-210). Measures were taken from both groups prior to the intervention, and every year thereafter for three years. Measures included data on basic demographics, functional autonomy, desire to be institutionalized, caregiver burden, and service utilization. Results did not indicate that ISD had any effect on mortality rates. Nevertheless, the study shows that ISD may contribute to increased quality of life, including less desire to be institutionalized, less deterioration, and decreased caregiver burden. There was no indication that ISD had an effect on utilization of services in either group.
A quasi-experimental study design was chosen because randomization was not possible in this context. The researchers took great care to select a comparable control environment to control any potential biases due to differences between the treatment and control groups. In comparison to the study environment, the control environment had similar percentages of people over 65 (11%), all who had similar access to services and utilization rates of these services. Furthermore, the rate of participants who refused to continue in the study (less than 5%) and the desire to be institutionalized (about 25%) was similar for both groups. Nevertheless, there were some important differences between the groups, which could pose a threat to the study’s internal validity. Groups differed in their score of functional autonomy, with the treatment group measuring as more autonomous than the control group. As the study proceeded, there was a significant loss of subjects - 72% remained for the first year, 54% for the second year, and 45% for the third and final year of the study - due to death and institutionalization. Though this was similar in both groups, thereby eliminating internal validity due to maturation, the cumulative attrition could have an effect on the study’s overall power. Lastly, because this study explored a specific program for a small cohort in only two communities, the generalizability of the results is limited.
Tourigny, A., Durand, P.J., Bonin, L., Hebert, R., and Rochette, L. (2004). Quasi-experimental study of the effectiveness of an integrated services delivery network for the frail elderly. Canadian Journal of Aging, 23(3), 231-246.