The sensitivity and specificity of computerised or paper and pencil cognitive assessments
Posted on 14 July 2014 in Healthcare
Improving timely dementia diagnosis is central to improving outcomes. In the UK, primary care is the gatekeeper to diagnostic services. Despite the importance of cognitive assessment in understanding the presentation of those with suspected dementia-related diseases, there is little standardisation of cognitive tests used across services and great variability in the sensitivity and specificity of these tests. Here, we present an economic model illustrating the effect of sensitivity and specificity of cognitive tests on the cost-effectiveness of the diagnostic pathway in the UK.
A decision tree model was developed to assess the cost-effectiveness of different cognitive assessments in primary care, compared to no systematic assessment. Sensitivity and specificity of cognitive assessments in detecting mild-moderate dementia was considered for Cantab Mobile , MMSE and the GPCOG. All patients referred to diagnostic services were subjected to a standardised assessment following current NICE clinical guidelines (2006). The likelihood of presenting with subjective memory complaints (SMC) was calculated based on prevalence of conditions and rates of cognitive complaints. Input parameters were derived from published literature and cost sources (NHS Reference Costs 2012-13; PSSRU, 2013). One-way sensitivity analysis was conducted on key model parameters.
Systematic use of cognitive assessment in patients presenting with SMC in primary care resulted in cost savings, and the magnitude of cost saving reflected the sensitivity and specificity of the test used. Cantab Mobile (estimated sensitivity of 100% and specificity of 92% to detect mild to moderate Alzheimer’s disease) was estimated to save £165 per patient presenting to primary care. Compared to this, using the MMSE (sensitivity 88% and specificity of 86%; Lin et al., 2013), or GPCOG (85% sensitivity and 86% specificity; Brodaty et al., 2002), saved £158 per patient, reducing the cost-savings relative to Cantab Mobile by about 4% per patient presenting to primary care with SMC. Consistent assessment of those presenting with SMC using Cantab Mobile was estimated to save an additional £700 per cohort of 100 patients, compared to the MMSE and GPCOG.
The quality of cognitive assessment in primary care has significant implications for the costs of dementia diagnosis. Using a more sensitive and specific cognitive assessment such as Cantab Mobile has the potential to be highly cost-effective, by better supporting appropriate decision making and onward referral.
View the partnering paper 'Modelling the economic impact of CANTAB use in UK primary care in the dementia diagnostic pathway'