Complexity of inferences in polytree-shaped semi-qualitative probabilistic networks

Semi-qualitative probabilistic networks (SQPNs) merge two important graphical model formalisms: Bayesian networks and qualitative probabilistic networks. They provade a very Complexity of inferences in polytree-shaped semi-qualitative probabilistic networks and qualitative probabilistic networks. They provide a very general modeling framework by allowing the combination of numeric and qualitative assessments over a discrete domain, and can be compactly encoded by exploiting the same factorization of joint probability distributions that are behind the bayesian networks. This paper explores the computational complexity of semi-qualitative probabilistic networks, and takes the polytree-shaped networks as its main target. We show that the inference problem is coNP-Complete for binary polytrees with multiple observed nodes. We also show that interferences can be performed in time linear in the number of nodes if there is a single observed node. Because our proof is construtive, we obtain an efficient linear time algorithm for SQPNs under such assumptions. To the best of our knowledge, this is the first exact polynominal-time algorithm for SQPn. Together these results provide a clear picture of the inferential complexity in polytree-shaped SQPNs.

Medication adherence in frail elderly outpatients with dementia

Since drug therapy in the elderly is complex and longterm and aged people commonly present some level of impairment and disability, medication adherence tend to decrease with age. Cognitive function is a key factor associated with medication adherence and professional or caregiver assistance may be necessary to maintain correct drug use. This study aims to analyze frail elderly outpatients aged 80 years or over diagnosed with dementia. The study is cross-sectional and is being conducted at the Ambulatory of Frailty of the University Hospital of the University of São Paulo (AF-UH). It is being based on information collected through an interview conducted with the patient or its caregiver. Medication adherence is assessed by the proportion of the prescribed drugs used in concordance with the prescription. Here it is presented the results of a pilot study. Thirty patients were included in the pilot study of which 23 (76.7%) were female and 7 (23.3%) males. The mean(SD) age, number of dwelling relatives, living children and prescribed drugs was, respectively, 86(5) years, 3(2), 3(2) and 6(3). The AF-UH consultation is the only regular physician encounter for 60.7% of the patients. Out of 30 patients, 5 (16.7%) live alone. Medication is a caregiver responsibility in 22 (73.4%) patients; the others (26.6%) self-administer their medicines. 13 (43.3%) of patients regularly use at least one drug not prescribed. Dementia was present in 8 patients all of which have a caregiver responsible for the management and,or the administration of the medicines; on the other hand, only 4 of the 22 nondemented patients (18.2%) have assistance of a caregiver (p<.001). The mean(SD) number of prescribed drugs was higher in nondemented patients [6.5(2.4)] than in those with dementia[3.5(2.3)] (p=.004). Educational level was similar between caregivers and patients (p=.503) as well as between caregivers of demented and non demented patients (p=.582). Among patients without dementia, those with caregiver assistance pre-presented the same mean(SD) medication adherence [0.93(0.14)] than those without it [0.78(0.28)] (p=.305). When compared to nondemented patients without caregivers, demented patients showed higher medication adherence [1.00(0.00)] (p=.013) since all of them used their drugs as recommended. The lower number of prescribed drugs and caregiver assistance seem to play an important role in the adherence of pharmacotherapy of demented patients in the studied population.

Medication reconciliation and adherence: a call for clinical pharmacist

Introduction and Objectives: With the population ageing, there is a growing number of people who have several comorbidities and make use of a variety of drugs. These factors lead to a greater predisposition to adverse drug events, as well as to medication errors. The clinical pharmacist is the most indicated health professional to target these issues. The aims of this study were to analyze the profile of medication reconciliation and assess the role of the clinical pharmacist regarding medication adherence. Material and Methods: Prospective observational cohort study conducted from Jan-Mar 2013 at the Surgical Clinic of the University Hospital of the University of Sao Paulo. 117 admitted patients - over the age of 18 years, under continuous medication use and with length of hospitalization up to 120h - were included. Discrepancies were classified as intentional/unintentional and according to their risk to cause harm, and interventions were divided into accepted/not accepted. Medication adherence was measured by Morisky questionnaire. Results and Conclusions: Only 30% of hospital prescriptions showed no discrepancies between the medications that the patient was using at home and those which were being prescribed at the hospital and more than one third of those had the potential to cause moderate discomfort or clinical deterioration. One third of total discrepancies were classified as unintentional. About 90% of the interventions were accepted by the medical staff. In addition, about 63% of patients had poor adherence to drug therapy. The study revealed the importance of the medication reconciliation at patient admission, ensuring greater safety and therapeutic efficacy of the treatment during hospitalization, and orienting the patient at discharge, assuring the therapy safety.

CID: an efficient complexity-invariant distance for time series

The ubiquity of time series data across almost all human endeavors has produced a great interest in time series data mining in the last decade. While dozens of classification algorithms have been applied to time series, recent empirical evidence strongly suggests that simple nearest neighbor classification is exceptionally difficult to beat. The choice of distance measure used by the nearest neighbor algorithm is important, and depends on the invariances required by the domain. For example, motion capture data typically requires invariance to warping, and cardiology data requires invariance to the baseline (the mean value). Similarly, recent work suggests that for time series clustering, the choice of clustering algorithm is much less important than the choice of distance measure used.In this work we make a somewhat surprising claim. There is an invariance that the community seems to have missed, complexity invariance. Intuitively, the problem is that in many domains the different classes may have different complexities, and pairs of complex objects, even those which subjectively may seem very similar to the human eye, tend to be further apart under current distance measures than pairs of simple objects. This fact introduces errors in nearest neighbor classification, where some complex objects may be incorrectly assigned to a simpler class. Similarly, for clustering this effect can introduce errors by “suggesting” to the clustering algorithm that subjectively similar, but complex objects belong in a sparser and larger diameter cluster than is truly warranted.We introduce the first complexity-invariant distance measure for time series, and show that it generally produces significant improvements in classification and clustering accuracy. We further show that this improvement does not compromise efficiency, since we can lower bound the measure and use a modification of triangular inequality, thus making use of most existing indexing and data mining algorithms. We evaluate our ideas with the largest and most comprehensive set of time series mining experiments ever attempted in a single work, and show that complexity-invariant distance measures can produce improvements in classification and clustering in the vast majority of cases.