Blood glucose and stroke.

OPINION STATEMENT: Current recommendations for the treatment of abnormal blood glucose levels for stroke rely on evidence-based guidelines. Ongoing advances in neuroimaging, tissue studies, animal studies, and case series and reports present findings that expose additional variables to be considered in the causal analysis of the role of diabetes and glucose control for stroke occurrence and outcome. The physician when treating the individual patient must fold this information into that provided by large clinical trials. Consideration of all information available without limitation to results from large, double-blind randomized trials allows the physician to expect and thus forecast the result of clinical action by understanding the complexity of the situation. Avoiding hypoglycemia and levels of hyperglycemia able to produce symptoms of stroke and cerebral damage by themselves would seem common sense. The plethora of evidence consistent with worse outcome and increased mortality in acutely ill and stroke patients with hyperglycemia also would call for treatment, with prudent avoidance of hypoglycemia. These rules appear to apply for hemorrhage as well as ischemia, and aggressive recommendations for lowing blood glucose caution against hypoglycemia.

Principled versus statistical thinking in diagnosis and treatment of stroke.

Medical science is now synonymous with probability-based statistics. Statistics deals with a group; it does not need probability theory. Probability theory is consistent with the worldview that the universe is infinite, bounded, random, and governed by chance. Its logic is binary, its geometry is Cartesian, its rules offer a scientific method by which hypotheses may be tested. Clinical trials and even hypothesis testing at the bedside have nestled into the probability foundation. As a result, scientific "evidence" now appears only through the lens of probability theory. Because there is no definitive truth in the worldview of probability theory, the truth of evidence lies in probabilities only. The probabilistic view of science has a firm impact on the practice of medicine and implications for medical-legal decisions.

Influence of cyclooxygenase-1 genotype on ex vivo aspirin response in patients at risk for stroke.

BACKGROUND: We sought to determine whether cyclooxygenase-1 (PTGS1) genotype is associated with the ability of aspirin to inhibit platelet aggregation in patients at risk for stroke. METHODS: Blood and urine samples were collected from 60 subjects, including 28 African Americans, who were taking aspirin for primary or secondary stroke prevention. Samples were analyzed for the PTGS1 A-707G, PTGS1 P17L, and glycoprotein IIIa (ITGB3)P1(A1/A2) genotypes, ex-vivo platelet aggregation, serum cholesterol, plasma salicylate levels, and urinary 11-dehydrothromboxane B(2) (11-dhTxB(2)) concentrations. The association between PTGS1 A-707G and P17L genotypes and aspirin response, as assessed by ex vivo studies and 11-dhTxB(2) concentrations, was evaluated by statistical testing and nonlinear mapping. RESULTS: Salicylate concentrations, ITGB3 genotype distribution and 11-dhTxB(2) concentrations were similar among PTGS1 genotype groups. More subjects with the PTGS1 17PP versus PL genotype had incomplete ex-vivo inhibition of platelet aggregation by aspirin (57 vs. 20%; p = 0.04). Fifty-nine percent of subjects homozygous for both the PTGS -707A and 17P alleles, but none with both the PTGS1 -707G and 17L alleles had incomplete inhibition with aspirin; p = 0.04. Similarly, nonlinear mapping showed a direct relationship between the PTGS1 17P allele and decreased aspirin response. When analyzed separately by ethnicity, the association with the P17L genotype and aspirin response persisted in African Americans, but not Caucasians. CONCLUSIONS: Our data suggest that the PTGS1 P17L genotype contributes to response to aspirin as assessed by ex-vivo platelet aggregation. Our data further suggest that the association between PTGS1 genotype and aspirin response might vary by ethnicity.

Platelet aggregation and recruitment with aspirin-clopidogrel therapy.

BACKGROUND: Aspirin-clopidogrel combination therapy inhibits platelet aggregation. The effect on platelet recruitment is unknown. METHODS: Thirty chronic ischemic stroke patients taking aspirin alone followed by aspirin-clopidogrel combined therapy had platelet reactivity tests performed over 3 months: ex vivo platelet aggregation, platelet recruitment and urinary 11-dehydro-thromboxane B(2) (11-dhTxB(2))excretion. Statistical analysis of variance compared platelet aggregation and recruitment between aspirin alone and aspirin-clopidogrel, and longitudinal regression analysis estimated platelet recruitment over time. Nonlinear mapping defined variable connections in each patient. RESULTS: Statistically significant differences were found between aspirin alone and aspirin-clopidogrel for (1) adenosine-diphosphate- and collagen-induced platelet aggregation and maximum inhibition of platelet recruitment and (2) increasing inhibition of platelet recruitment over time. Urinary 11-dhTxB(2) excretion did not predict platelet aggregation response. Nonlinear mapping showed patient-unique variable interconnections. CONCLUSIONS: Platelet inhibition with aspirin-clopidogrel may increase over time, and future studies should focus on this finding in the context of vascular complications.

Carotid endarterectomy for asymptomatic plaque.

Statistical correlations are linear noninteractive relationships, but the dynamics of causation are nonlinear and involve complex interactions where variables change through their effect on one another and interact with the context of the patient over time. The discovery and interpretation of plaque vulnerable features in the individual patient are not determined for the asymptomatic patient being considered for carotid endarterectomy. New technologies for identification of plaque chemical and morphologic composition are on the horizon and may be applicable to certain patients but change in their usefulness as the plaque and patient change over time.

From fuzzy logic toward plurimonism: the science of active and empathic observation.

Plurimonism is a new philosophy and method of science. It holds that the revolution in computer science and artificial intelligence has reached the point that all the sciences in general can now account for the complex relations of an irreducible plurality of unique observers engaged in describing the same event. Plurimonism seeks to describe the conscious and unconscious relations of the scientific observer during the act of observation of a given event while preserving the historical uniqueness and indivisible identity of each such observer. Using the framework of plurimonism, we mathematically formulate the problem of empathy. This self-reflective mathematical model entails four components of the empathic process involving two observers. They are: 1) the self; 2) the self's-other; 3) the other; and 4) the other's-self. It measures the degree of accuracy of the therapist-observer's empathy, as well as conscious and unconscious processes involved in the patient-observer's idealization and the therapist-observer's confidence in clinical psychotherapy. Ratings are obtained from both patient and therapist from four different points of view. The plural views of the patient's global assessment of functioning (GAF) are from: 1) the therapist's view (TGAF); 2) the patient's view (PGAF); 3) the therapist empathic view (TEGAF), which represents the therapist's estimate of PGAF; and 4) the patient's empathic estimate of the TGAF. The GAF scale is the standard dimensional 100-point-scale measure used in psychiatry for recording a patient's functioning. The patient's estimate of the therapist's degree of accuracy as well as the therapist's confidence in his or her empathic accuracy is also represented. Three formulae are presented that describe the degree of the therapist's empathic accuracy, the patient's over-idealization/under-idealization, and the therapist's over-confidence/under-confidence. The concept of empathy is here restricted to mean the degree to which one observer can take the point of view of another observer when both are observing the same thing.

Measurable differences between sequential and parallel diagnostic decision processes for determining stroke subtype: a representation of interacting pathologies.

Stroke diagnosis depends on causal subtype. The accepted classification procedure is a succession of diagnostic tests administered in an order based on prior reported frequencies of the subtypes. The first positive test result completely determines diagnosis. An alternative approach tests multiple concomitant diagnostic hypotheses in parallel. This method permits multiple simultaneous pathologies in the patient. These two diagnostic procedures can be compared by novel numeric criteria presented here. Thrombosis, a type of ischemic stroke, results from interaction between endothelium, blood flow and blood components. We tested for ischemic stroke on thirty patients using both methods. For each patient the procedure produced an assessment of severity as an ordered set of three numbers in the interval [0, 1]. We measured the difference in diagnosis between the sequential and parallel diagnostic algorithms. The computations reveal systematic differences: The sequential procedure tends to under-diagnose and excludes any measure of interaction between pathologic elements.

Evolving concepts in the treatment of venous thromboembolism: the role of factor Xa inhibitors.

Anticoagulation is an essential component of the care of patients with venous thromboembolism (VTE). Traditional anticoagulants for the treatment of VTE include unfractionated heparin (UFH), low-molecular-weight heparin (LMWH), and the oral vitamin K antagonist, warfarin. A variety of anticoagulant agents with improved pharmacologic and clinical profiles are emerging and offer benefits over the traditional therapies. One of the most recent advances has been the development of new agents, such as oral direct thrombin inhibitors and factor Xa inhibitors, that have a more selective and targeted effect on the coagulation cascade. Recent clinical trials have evaluated fondaparinux, the first commercially available factor Xa inhibitor, in the treatment of patients with deep vein thrombosis and pulmonary embolism and indicate efficacy and safety as compared with traditional options such as UFH and LMWH. Fondaparinux is a welcomed addition to the available antithrombotic options.

The application of fuzzy logic to the prescription of antithrombotic agents in the elderly.

The prescription of antithrombotic agents in the elderly depends, to certain degree, on the identity of the unique individual elderly patient. This dependence cannot be captured by viewing the patient as belonging to a group, but rather by viewing age in the context of unique individual biology. This context is historical, physiological, psychological, and time- and location-dependent. The group-based approach to patient therapy is found in evidence-based medicine, typified by the large double-blind randomised clinical trial from which clinical recommendations are defined. An alternative approach to capturing the unique context of each patient is based using fuzzy logic and mathematics. In particular, the fuzzy subsethood theorem of Kosko has direct application here. A new measure of clinical efficiency, K, derived from the fuzzy subsethood theorem has redefined the clinical significance of age. In particular, the causal role of age in any one patient's response to therapy is to unique degree for that patient. This is because the causal measure K accounts for the role of known and unknown contextual factors of the patient, most of which are unknown, in defining clinical effect in any specific patient. Thus, we have shown mathematically why 'age', or being 'elderly', is only one factor to be taken into consideration when therapeutic decisions regarding antithrombotic therapy are made in any given patient. This is contrary to the group-based approach of evidence-based medicine, where age has the same therapeutic significance for all patients. This is because any hypothesis of evidence-based medicine is group-based and cannot be extrapolated to the individual patient. Such extrapolation has to be personalized through the expertise of the physician. The physician takes into account all the factors not considered in any therapeutic group-based trial which apply to his specific patient. These considerations take into account past experience with that patient. An example of the effect of age on the patient and his/her therapeutic context is provided which shows how age affects different patients to different degree using the fuzzy causal measure K. The purpose of this exercise is to show that there is mathematical support for the argument that individualization of patient therapy by expert decision has measurable clinical significance. What is different here is that measure stick is not probabilistic but fuzzy-mathematic based.

Warfarin in patients with stroke and reasons for discontinuation.

BACKGROUND: Reasons for warfarin prescription in patients who have had ischemic stroke, and its discontinuation, are complex and unique for each individual patient. OBJECTIVE: We sought to discover the reasons for discontinuation of warfarin therapy in patients followed up in a medical center antithrombosis clinic after stroke occurrence. MATERIALS: In all, 229 patients on warfarin therapy with history of stroke were followed up in our antithrombosis clinic between January 1997 and March 2003. Of these patients, 132 were identified as having left the medical center antithrombosis clinic. Reasons for discontinuation of therapy were noted. Patients on combination antiplatelet-warfarin therapy were identified, as were the reasons for antithrombotic therapy. RESULTS: The most common reason for discontinuation of warfarin was noncompliance with therapy and failure to show up for scheduled visits to the antithrombosis clinic. Hemorrhage was rare and not solely attributed to anticoagulation alone. The complication rate for major hemorrhage was 2.18%, well within the accepted range for US antithrombosis clinics. This rate included those on combination antiplatelet-anticoagulant therapy. CONCLUSIONS: Combination therapy of warfarin and antiplatelet agents may not be dangerous in this group of patients. Careful assessment of patients on warfarin therapy in a medical center-based antithrombosis clinic assures close supervision of compliance, potential complications, and general medical and environmental conditions of the patient and allows for a controlled discontinuation of anticoagulation if necessary.

Artificial adaptive systems and fuzzy measures for translation of clinical trial results to the bedside.

Medical science relies uniquely on statistical evidence from large clinical trials or laboratory experiments to deal with uncertainties regarding clinical decisions. The statistical evidence is stated in probabilities. Probability theory is based on the logical rules set forth by Aristotle: the law of noncontradiction, excluded middle, and identity. Thus, medical science assumes that variables such as disease and treatment relate or interact only in certain ways and the statistical evidence defines the presence of these relations. Experiments seek to find these certain predefined relations among variables, relations that then predict in terms of probabilities their behavior over time. Living systems are nonlinear and complex, which means that the relations among variables not only change over time, but their behavior over time changes in response to the change in one another in unpredictable ways. A patient is a living system and as such is a dynamic system of the type described. Because patients have different genetic and contextual makeup, there is no reason to assume they share a common dynamic. One goal of medical science should be to discover the behavioral dynamic of variables of interest in a given patient to individualize treatment appropriate to that system. Ideally this would be in real time so that diagnostic and treatment decisions adapt to dynamic change. Artificial adaptive systems on the one hand and fuzzy theory on the other are two scientific approaches toward this end.

Measurable prediction for the single patient and the results of large double blind controlled randomized trials.

BACKGROUND: It has been shown that the clinical state of one patient can be represented by known measured variables of interest, each of which then form the element of a fuzzy set as point in the unit hypercube. We hypothesized that precise comparison of a single patient with the average patient of a large double blind controlled randomized study is possible using fuzzy theory. METHODS/PRINCIPLE FINDINGS: The sets as points unit hypercube geometry allows fuzzy subsethood to define in measures of fuzzy cardinality different conditions, similarity and comparison between fuzzy sets. A fuzzy measure of prediction is defined from fuzzy measures of similarity and comparison. It is a measure of the degree to which fuzzy set A is similar to fuzzy set B when different conditions are taken into account and removed from the comparison. When represented as a fuzzy set as point in the unit hypercube, a clinical patient can be compared to an average patient of a large group study in a precise manner. This comparison is expressed by the fuzzy prediction measure. This measure in itself is not a probability. Once thus precisely matched to the average patient of a large group study, risk reduction is calculated by multiplying the measured similarity of the clinical patient to the risk of the average trial patient. CONCLUSION/SIGNIFICANCE: Otherwise not precisely translatable to the single case, the result of group statistics can be applied to the single case through the use of fuzzy subsethood and measured in fuzzy cardinality. This measure is an alternative to a Bayesian or other probability based statistical approach.

The difference between a dynamic and mechanical approach to stroke treatment.

The current classification of stroke is based on causation, also called pathogenesis, and relies on binary logic faithful to the Aristotelian tradition. Accordingly, a pathology is or is not the cause of the stroke, is considered independent of others, and is the target for treatment. It is the subject for large double-blind randomized clinical therapeutic trials. The scientific view behind clinical trials is the fundamental concept that information is statistical, and causation is determined by probabilities. Therefore, the cause and effect relation will be determined by probability-theory-based statistics. This is the basis of evidence-based medicine, which calls for the results of such trials to be the basis for physician decisions regarding diagnosis and treatment. However, there are problems with the methodology behind evidence-based medicine. Calculations using probability-theory-based statistics regarding cause and effect are performed within an automatic system where there are known inputs and outputs. This method of research provides a framework of certainty with no surprise elements or outcomes. However, it is not a system or method that will come up with previously unknown variables, concepts, or universal principles; it is not a method that will give a new outcome; and it is not a method that allows for creativity, expertise, or new insight for problem solving.

Factors influencing warfarin dose requirements in African-Americans.

INTRODUCTION: African-Americans are under-represented in studies assessing contributors to warfarin response. Our primary objective was to determine whether the genes for cytochrome P450 (CYP) 2C9, nicotinamide adenine dinucleotide phosphate, reduced, quinone oxidoreductase (NQO1) and vitamin K epoxide reductase complex subunit 1 (VKORC1) are associated with warfarin dose requirements in African-Americans. PATIENTS AND METHODS: The following factors were assessed: demographics; clinical data; the CYP2C9 Arg144Cys (*2), Ile358Leu (*3) and Asp360Glu (*5); NQO1 Pro187Ser (*1/*2); and VKORC1 G6853C genotypes were analyzed in 115 African-Americans on stable warfarin doses. RESULTS: Allele frequencies were 0.05 for the CYP2C9 *2, *3 or *5 alleles; 0.20 for NQO1 *2; and 0.25 for VKORC1 6853C. Possession of a CYP2C9*2, *3 or *5 allele was associated with a 38% lower warfarin dose compared with the *1/*1 genotype (30 +/- 13 vs 48 +/- 18 mg/week; p = 0.003). Neither the NQO1 *1/*2 nor VKORC1 G6853C genotype was associated with warfarin dose requirements in the population as a whole or in CYP2C9*1 allele homozygotes. Multiple regression analysis revealed that CYP2C9 genotype (p = 0.015), age (p < 0.001) and body surface area (p < 0.001) were jointly associated with warfarin dose requirements, and together explained 33% of the variability in warfarin dose requirements among African-Americans. DISCUSSION: Our data suggest that CYP2C9 genotype, age and body size are important determinants of warfarin dose requirements in African-Americans. Our data further suggest that the VKORC1 G6853C polymorphism alone may not be useful for predicting warfarin dose requirements in this racial group.

Thrombophilia in ischemic stroke subtypes: implications for treatment.

The current understanding of thrombogenesis is modeled on Virchow's triad: stasis, hypercoagulability, and vessel wall injury. There is a dynamic (always changing) nonlinear interaction between the vascular wall, blood components, and flow, which at times defined "pathologic" leads to thrombosis or hemorrhage, at other times called "healthy" to normal hemostasis. The triad named after Virchow was not designated as such in Virchow's work. Instead, Virchow showed that thrombosis itself leads to endothelial damage, hypercoagulability, and stasis. Thus, cause and effect regarding the elements of Virchow's triad and thrombosis become indistinguishable if linearity is considered mandatory. Considering a nonlinear relation solves this problem. In the real patient, each element is present to a degree. At every moment in time, the direction of coagulation (toward hemostasis, thrombosis, or hemorrhage) and the dynamic of interaction of the elements of the triad change. The complexity and nonlinearity of the thrombotic context is evident. These facts suggest a new venue for diagnostic classification of stroke (ischemic and hemorrhagic) by causation and have implications for its prevention and treatment. Clinical and laboratory evidence can be gathered for the elements of Virchow's triad as well as for fibrinolysis and thrombosis. Mathematical methods other than probability-based statistics can represent the measured presence of these elements to a degree and their nonlinear relationship. These include, but may not be limited to, Riemannian geometry, fuzzy logic, cellular automata, and infinitesimals, all proscribed by evidence-based medicine. However, by using these methods, diagnosis and treatment measures for stroke can be built on a causal rather than risk methodology, individualizing medical decisions to the patient. All current clinical guidelines are based on linear methods of probability-based statistics and group-based data. The therapeutic choice of antithrombotic therapy in the individual patient for whom measured elements of thrombogenesis are available rests on the knowledge and expertise of the treating physician.

Sex difference in the antiplatelet effect of aspirin in patients with stroke.

BACKGROUND: There is substantial interpatient variability in response to aspirin after an ischemic stroke or transient ischemic attack (TIA), as assessed by ex vivo effects of aspirin on platelet aggregation. The factors contributing to this variability are not well defined. OBJECTIVE: To determine whether demographic, social, or clinical characteristics are associated with ex vivo response to aspirin in patients with a history of stroke or TIA. METHODS: Eighty-one patients who were taking aspirin for secondary stroke prevention and underwent ex vivo platelet aggregation studies were identified. The medical records of eligible patients were reviewed by clinicians who specialize in the management of stroke patients. Characteristics were compared between 45 patients who had a complete response to aspirin and 36 patients who exhibited an incomplete (partial) response to aspirin based on the results of platelet aggregation testing. RESULTS: The median (range) aspirin dose was similar in complete (325; 81-1950 mg/day) and partial (325; 81-1300 mg/day) responders. There was no association between aspirin response and age, race, body mass index, medical history, smoking status, or use of statin or hormone replacement therapy. However, sex was significantly associated with response to aspirin, with more women in the partial versus complete responder group (75% vs 49%; p = 0.02). CONCLUSIONS: Our data suggest that aspirin may be less effective at inhibiting platelet aggregation in women compared with men who have a history of ischemic stroke or TIA.

Fuzzy logic and continuous cellular automata in warfarin dosing of stroke patients.

"Evidence-based" recommendations for warfarin prescription in patients with history of ischemic stroke limit its use to prevention of stroke due to atrial fibrillation. Warfarin is also prescribed by the authors to prevent thrombosis in stroke patients with thrombophilia and potential cardiac or arterial source for thromboembolism. These potential conditions, in the face of thrombophilia, include, but may not be limited to, dilated cardiomyopathy, decreased left ventricular function, atrial septal aneurysm with or without patent foramen ovale (PFO), PFO with evidence of pelvic or lower extremity deep venous thrombosis or with clear thrombophilia, spontaneous echocardiographic contrast, intracardiac or intra-arterial thrombus, intra-aortic arch thrombus, high degree of stenosis of large- and medium-sized cerebrovascular arteries, and arterial dissection. Commonly diagnosed thrombophilic states in our population currently include protein S or C deficiency, antiphospholipid antibodies, and less commonly ATIII deficiency, factor V Leiden mutation, G20210A PT mutation, and plasminogen activator inhibitor-1 mutation. Thrombophilic states often occur in combination. The occurrence of combined arterial, cardiac, and thrombophilic sources of thromboembolism poignantly describes the complexity of causation of ischemic stroke in any one patient. Our practice of treating the complex interaction of thromboembolic sources is based on scientific evidence, which is not arbitrarily limited to probability-based statistics. Warfarin is well known in the clinical setting to interact with many different contextual factors of the individual patient, making its dosing and response unique to that patient. We have shown why the indications for warfarin use and its dosing cannot be directly extrapolated to the individual patient from the results of large, double-blind, randomized trials. In practice, the unique patient and his or her context must be considered by the expert physician who makes the therapeutic decision. The context includes, but is not limited to, known pathologies that contribute to thrombus formation according to the accepted pathophysiologic model of thrombosis based on Virchow's triad of altered flow, endothelium, and blood components.

Concomitant drug, dietary, and lifestyle issues in patients with atrial fibrillation receiving anticoagulation therapy for stroke prophylaxis.

Atrial fibrillation is a common cardiac arrhythmia and the leading risk factor for stroke. In those at moderate to high risk of stroke, oral anticoagulation therapy with warfarin (a vitamin K antagonist) significantly reduces not only the frequency of such events but also their severity and the associated risk of death. However, achieving optimal anticoagulation with this agent is clinically challenging in view of its complex pharmacokinetic and pharmacodynamic profile. In this regard, concomitant drug therapy (both prescription and over-the-counter medications, including herbal products, vitamins, and various nutritional supplements), along with alcohol intake, dietary factors, and changes in lifestyle, can significantly affect anticoagulation control and thereby expose patients to the risk of bleeding or thromboembolic complications (due to over- and underanticoagulation, respectively). Therefore, it is recommended that intensified monitoring of anticoagulation be performed at initiation and discontinuation of concomitant drug therapy, and in the case of significant dietary and lifestyle changes. Moreover, many patients receive inadequate education and are unaware of such risks and their implications, highlighting the need for better awareness and education on this important aspect of anticoagulation therapy.

The selection of antithrombotic agents in the prevention of recurrent ischemic stroke.

The scientific selection of antithrombotic therapy has been dominated by group-based interpretation of data in the form of probability-based statistics in evidence-based medicine. Because the data in large randomized trials are grouped and averaged, the relationship to initial conditions of the patient is lost. There is a pathologic model and basis by which antithrombotic agents may be chosen for prevention of recurrent thrombus and thromboembolism in patients with stroke. This model applies in all settings, but has not been tested when the elements of the model remain connected to the individual patient and his or her unique context. Alternative math models and perception-based science respect the criticality of initial conditions and capture the rules that apply to the actual causal mechanisms within the patient's body. Individualized patient rather than group-based choices insure thrombus-type specific targeted therapy.

Fuzzy logic and causal reasoning with an 'n' of 1 for diagnosis and treatment of the stroke patient.

The current scientific model for clinical decision-making is founded on binary or Aristotelian logic, classical set theory and probability-based statistics. Evidence-based medicine has been established as the basis for clinical recommendations. There is a problem with this scientific model when the physician must diagnose and treat the individual patient. The problem is a paradox, which is that the scientific model of evidence-based medicine is based upon a hypothesis aimed at the group and therefore, any conclusions cannot be extrapolated but to a degree to the individual patient. This extrapolation is dependent upon the expertise of the physician. A fuzzy logic multivalued-based scientific model allows this expertise to be numerically represented and solves the clinical paradox of evidence-based medicine.