Alzheimer hastalığı için eniyi tarama programları

Alzheimer's Disease (AD) is believed to be the most common type of dementia. Even though screening for AD has been discussed widely, there is no screening pro- gram integrated as a policy in any country. To the best of our knowledge this work is the first one that investigates optimal screening policies for AD using Markov models. In our first model we use Mini-Mental State Examination (MMSE) as the screening tool in line with current clinical practice. The reward function takes into consideration quality-adjusted life years (QALYs) and treatment costs. It is a fully observable model which assumes that the true state of an individual can be observed with certainty at any time. We use Alzheimer's Disease Neuroimaging Initiative (ADNI) data to compute state-to-state transition probabilities and incorporated QALY and cost data from the literature. We find that the optimal decision is not to screen the population using MMSE under current levels of treatment effectiveness. Parameterization on this value results in changes of the optimal screening policy. These results and current medical research initiatives motivate us to build a more elaborate model which focuses on the preclinical stages of the disease as defined by the National Institute on Aging and Alzheimer's Association in 2011. In the new model, we introduce disease free and preclinical states as partially observable states, and the screening decision is taken while an individual is in one of the partially observable states. An additional completely observable treatment for preclinical state has been integrated along with mild cognitive impairment, AD and death states. This model considers cerebrospinal fluid (CSF) collection as a screening test. A partially observable MDP model provides modeling flexibility by relaxing the assumption that the true state of the patient can be observed. The output of our model is a screening program that specifies at what points in time an individual over 50 years of age with a given risk of AD will be directed to undergo screening in a way to maximize total expected QALYs. The transition probabilities between the states are computed with the support of the Knight Alzheimer's Disease Research Center database of Washington University in St. Louis. We analyzed the impact of the hypothetical treatment effectiveness and QALY of the corresponding state on the outcome. In addition, we computed the total expected discounted costs of implementing the optimal policies for individuals of different risk profile and showed that for a vast majority of instances, implementing screening policies are either cost-saving or cost-effective. Finally, we change the objective function by adding the related costs to obtain a societal perspective on the screening problem. Computing the costs, we see that implementing screening policies can save a significant proportion of the current estimated AD burden.