I assessed the accuracy of two clinical prediction rules, the d-dimer blood test and point of care ultrasound for diagnosing lower limb deep vein thrombosis. Emergency physicians were trained in ultrasound and prospectively scanned emergency department patients with suspected deep vein thrombosis. Accuracy of the Wells and AMUSE rules and the ultrasound result was compared to radiology-performed ultrasound and a 90-day clinical outcome. Univariate and multivariate analyses were performed assessing which factors were associated with the outcome. The sensitivity and specificity of the Wells score for the clinical outcome was 85.7% and 68.5%; the AMUSE score 85.7% and 54.4%. Ultrasound had a sensitivity of 91.7% and specificity of 91.7% for radiology-diagnosed thrombus and 78.6% and 95.0% for clinical outcome. The odds ratio of a positive outcome with a positive ultrasound was 65.1. After receiving the ultrasound training program, emergency physicians were unable to demonstrate sufficient accuracy to replace current diagnostic strategies.

Study Design: A blinded, prospective diagnostic test study was conducted.<p>Objectives: To assess the inter-tester reliability of clinical examination items for carpal tunnel syndrome (CTS), to assess the validity of individual clinical test items and self-report measures for the diagnosis of CTS and to assess the accuracy of an optimum test item cluster for the diagnosis of CTS.<p>Study Rational: Examination of the diagnostic validity of various clinical tests for the diagnosis of CTS has shown mixed results and the reliability of many of these tests has not been determined. The majority of the diagnostic validity research for CTS has examined tests individually which is in contrast to clinical practice where the results of multiple tests are combined as part of the clinical reasoning process in order to formulate a differential diagnosis: the test item cluster, derived through logistic regression, is proposed as a means to integrate the validity of multiple tests. Methods: 37 subjects (74 hands) were recruited from a convenience sample of consecutive patients referred to for electrodiagnostic (EDX) testing due to upper extremity symptomatic complaints. Subjects underwent EDX testing followed by completion of self-report questionnaires and a standardized clinical examination by examiners blinded to EDX results. Diagnostic validity was determined for both general and restricted CTS classification groupings. <p>Results: Out of 18 clinical test items, 12 had reliability coefficients (i.e. ICC or Kappa) of .40 or greater. There were 10 clinical exam and self-report items that were found to have likelihood point estimates above 2 or below 0.50 for the general diagnostic classification and 6 items had acceptable validity for the restricted classification. The test item cluster (TIC) derived for the general classification included hand numbness and symptom reproduction with the upper limb neurodynamic test 1. The TIC derived for the restricted classification included hand numbness and the overall score of the symptom component of the Bringham Carpal Tunnel Questionnaire. The 95% confidence intervals for most likelihood ratio point estimates were wide. <p>Conclusions: The TICs for both classification groupings did not yield improved diagnostic validity beyond that found with the single best test item (hand numbness). The value of the single best test item hand numbness was in a negative response. Further investigation is required to validate the TIC and the single best test item and to improve the point estimate precision.

BACKGROUND: MRI is increasingly being used for rectal cancer staging. The purpose of this study was to summarize published evidence to determine the accuracy of MRI for T-category, lymph node (LN) metastases, and circumferential resection margin (CRM) involvement in rectal cancer. METHODS: Sensitivity, specificity, and diagnostic odds ratios (DOR) were estimated using hierarchical summary receiver operating characteristics modeling and bivariate random effects modeling. RESULTS: MRI was more specific for CRM (94%, 95%CI 88-97) than for T-category (75%, 95%CI 68-80) and LN’s (71%, 95%CI 59-81) but was more sensitive for T-category (87%, 95%CI 81-92) than for CRM (77%, 95%CI 57-90) and LN’s (77%, 95%CI 69-84). DOR was higher for CRM (56.1, 95%CI 15.3-205.8) than for LN’s (8.3, 95%CI 4.6-14.7) and T-category (20.4, 95%CI 11.1-37.3). CONCLUSIONS: MRI has good accuracy for both CRM and T-category and should be considered for preoperative rectal cancer staging. In contrast, LN assessment is poor on MRI.

BACKGROUND: MRI is increasingly being used for rectal cancer staging. The purpose of this study was to summarize published evidence to determine the accuracy of MRI for T-category, lymph node (LN) metastases, and circumferential resection margin (CRM) involvement in rectal cancer. METHODS: Sensitivity, specificity, and diagnostic odds ratios (DOR) were estimated using hierarchical summary receiver operating characteristics modeling and bivariate random effects modeling. RESULTS: MRI was more specific for CRM (94%, 95%CI 88-97) than for T-category (75%, 95%CI 68-80) and LN’s (71%, 95%CI 59-81) but was more sensitive for T-category (87%, 95%CI 81-92) than for CRM (77%, 95%CI 57-90) and LN’s (77%, 95%CI 69-84). DOR was higher for CRM (56.1, 95%CI 15.3-205.8) than for LN’s (8.3, 95%CI 4.6-14.7) and T-category (20.4, 95%CI 11.1-37.3). CONCLUSIONS: MRI has good accuracy for both CRM and T-category and should be considered for preoperative rectal cancer staging. In contrast, LN assessment is poor on MRI.

The modern medical diagnostic systems are based on the techniques using digital data formats – a natural feed for the computer based systems. With the use of modern diagnostic techniques the diagnosis process is becoming more complex as many diseases seem to have the same pre-symptoms at early stages. And of course computer based systems require more efficient and effective ways to identify such complexities. However, the existing formalisms for knowledge representation, tools and technologies, learning and reasoning strategies seem inadequate to create meaningful relationship among the entities of medical data i.e. diseases, symptoms and medicine etc. This inadequacy actually is due to the poor design of the knowledge base of the medical system and leads the medical systems towards inaccurate diagnosis. This thesis discusses the limitations and issues specific to the design factors of the knowledge base and suggests that instead of using the deficient approaches and tools for representing, learning and retrieving the accurate knowledge, use of semantic web tools and techniques should be adopted. Design by contract approach may be suitable for establishing the relationships between the diseases and symptoms. The relationship between diseases and symptoms and their invariants can be represented more meaningfully using semantic web. This can lead to more concrete diagnosis, by overcoming the deficiencies and limitations of traditional approaches and tools.

Study Design: A blinded, prospective diagnostic test study was conducted.<p>Objectives: To assess the inter-tester reliability of clinical examination items for carpal tunnel syndrome (CTS), to assess the validity of individual clinical test items and self-report measures for the diagnosis of CTS and to assess the accuracy of an optimum test item cluster for the diagnosis of CTS.<p>Study Rational: Examination of the diagnostic validity of various clinical tests for the diagnosis of CTS has shown mixed results and the reliability of many of these tests has not been determined. The majority of the diagnostic validity research for CTS has examined tests individually which is in contrast to clinical practice where the results of multiple tests are combined as part of the clinical reasoning process in order to formulate a differential diagnosis: the test item cluster, derived through logistic regression, is proposed as a means to integrate the validity of multiple tests. Methods: 37 subjects (74 hands) were recruited from a convenience sample of consecutive patients referred to for electrodiagnostic (EDX) testing due to upper extremity symptomatic complaints. Subjects underwent EDX testing followed by completion of self-report questionnaires and a standardized clinical examination by examiners blinded to EDX results. Diagnostic validity was determined for both general and restricted CTS classification groupings. <p>Results: Out of 18 clinical test items, 12 had reliability coefficients (i.e. ICC or Kappa) of .40 or greater. There were 10 clinical exam and self-report items that were found to have likelihood point estimates above 2 or below 0.50 for the general diagnostic classification and 6 items had acceptable validity for the restricted classification. The test item cluster (TIC) derived for the general classification included hand numbness and symptom reproduction with the upper limb neurodynamic test 1. The TIC derived for the restricted classification included hand numbness and the overall score of the symptom component of the Bringham Carpal Tunnel Questionnaire. The 95% confidence intervals for most likelihood ratio point estimates were wide. <p>Conclusions: The TICs for both classification groupings did not yield improved diagnostic validity beyond that found with the single best test item (hand numbness). The value of the single best test item hand numbness was in a negative response. Further investigation is required to validate the TIC and the single best test item and to improve the point estimate precision.

<p>The modern medical diagnostic systems are based on the techniques using digital data formats – a natural feed for the computer based systems. With the use of modern diagnostic techniques the diagnosis process is becoming more complex as many diseases seem to have the same pre-symptoms at early stages. And of course computer based systems require more efficient and effective ways to identify such complexities. However, the existing formalisms for knowledge representation, tools and technologies, learning and reasoning strategies seem inadequate to create meaningful relationship among the entities of medical data i.e. diseases, symptoms and medicine etc. This inadequacy actually is due to the poor design of the knowledge base of the medical system and leads the medical systems towards inaccurate diagnosis. This thesis discusses the limitations and issues specific to the design factors of the knowledge base and suggests that instead of using the deficient approaches and tools for representing, learning and retrieving the accurate knowledge, use of semantic web tools and techniques should be adopted. Design by contract approach may be suitable for establishing the relationships between the diseases and symptoms. The relationship between diseases and symptoms and their invariants can be represented more meaningfully using semantic web. This can lead to more concrete diagnosis, by overcoming the deficiencies and limitations of traditional approaches and tools.</p>

I assessed the accuracy of two clinical prediction rules, the d-dimer blood test and point of care ultrasound for diagnosing lower limb deep vein thrombosis. Emergency physicians were trained in ultrasound and prospectively scanned emergency department patients with suspected deep vein thrombosis. Accuracy of the Wells and AMUSE rules and the ultrasound result was compared to radiology-performed ultrasound and a 90-day clinical outcome. Univariate and multivariate analyses were performed assessing which factors were associated with the outcome. The sensitivity and specificity of the Wells score for the clinical outcome was 85.7% and 68.5%; the AMUSE score 85.7% and 54.4%. Ultrasound had a sensitivity of 91.7% and specificity of 91.7% for radiology-diagnosed thrombus and 78.6% and 95.0% for clinical outcome. The odds ratio of a positive outcome with a positive ultrasound was 65.1. After receiving the ultrasound training program, emergency physicians were unable to demonstrate sufficient accuracy to replace current diagnostic strategies.

A prevalência de diabetes mellitus (DM) está aumentando constantemente em todo o mundo a uma taxa alarmante. Devido às suas complicações que causam uma maior morbidade, invalidez e mortalidade, o DM representa um enorme problema para a saúde pública. Com isso, ações, tanto em diagnóstico como em tratamento, são necessárias para desacelerar a tendência atual e prevenir o desenvolvimento das complicações do DM. Recentemente, a hemoglobina glicada (HbA1c) foi introduzida nos critérios diagnósticos de DM. Os resultados de HbA1c são igualmente apropriados para o diagnóstico, apesar de não necessariamente detectarem DM nos mesmos indivíduos detectados pelos critérios de glicemia. No entanto, os níveis de HbA1c podem ser influenciados por qualquer condição que altere a vida útil dos eritrócitos e metabolismo da hemoglobina, independentemente da glicemia, resultando em erro diagnóstico neste grupo da população com estas condições. Além disso, estudos epidemiológicos revelaram que a glicemia pós-prandial tem um maior risco de causar complicações cardiovasculares em relação à hiperglicemia persistente e pode ser acessada com precisão usando a albumina glicada (AG) e não a HbA1c. Nesse contexto, estudos recentes têm evidenciado que a AG pode ser um marcador para diagnóstico do DM e também ser utilizado como um marcador alternativo à HbA1c para o controle glicêmico. No entanto, esses estudos, foram realizados apenas na população asiática e podem não ser aplicáveis a outros grupos étnicos. Por isso, mais investigações para a validação do desempenho diagnóstico da AG na predição do DM em diferentes grupos etnicos são necessárias. Neste estudo avaliamos o desempenho da AG no diagnóstico do DM em 242 indivíduos brasileiros com idade média de 54,4 anos (+ 13,0). Baseando-se nos valores de glicose plasmática durante o teste oral de tolerância à glicose (TOTG), o DM foi detectado em 31,8%. AG ≥16,8% apresentou acurácia similar para a detecção de DM conforme definido por HbA1c >6,5%. O uso da razão glicemia de 2h pós-sobrecarga de 75g de glicose (2hPG) e AG (2hPG/AG) melhora a sensibilidade, reduz o número de diagnósticos incorretos por AG ou HbA1c >6,5% e possui um acurácia comparável ao TOTG, indicando que o uso de uma estratégia aplicando a razão da glicemia pós-prandial (GPP) real e AG (GPP/AG) pode ser mais conveniente para pacientes e aumentar o desempenho diagnóstico do teste. Estudos para validar esta estratégia são necessários. / The prevalence of diabetes mellitus (DM) is constantly increasing worldwide at an alarming rate. Due to its complications that cause greater morbidity, disability and mortality, DM represents a heavy burden on public health. Therefore, actions in both, diagnosis and treatment, are necessary to slow down the current tendency and avoid the development of DM complications. Recently, the glycated hemoglobin test (HbA1c) was introduced in the diagnostic criteria for DM. The HbA1c results are equally appropriate for diagnostic testing, though not necessarily detect DM in the same subjects detected by plasma glucose criteria. However, blood HbA1c levels may be influenced by any condition that changes the lifespan of erythrocytes and hemoglobin metabolism regardless of glycemia, resulting in the misdiagnosis of this population group. In addition, epidemiological studies have shown that postprandial glycemia has a higher risk of causing cardiovascular complications than chronic hyperglycemia and can be accurately assessed using the glycated albumin (GA) test rather than HbA1c. In this context, recent studies have shown that GA may be a marker for the diagnosis of DM and also be used as an alternative marker to HbA1c on many occasions. However, these studies have been conducted only in the Asian population and may not be applicable to other ethnic groups. Therefore, further investigations to validate the diagnostic performance of GA in the prediction of DM in different ethnic groups are necessary. In this study, we evaluated the GA performance in the diagnosis of DM in 242 Brazilian individuals with a mean age of 54.4 years (+ 13.0). Based on plasma glucose values during oral glucose tolerance test (OGTT), DM was detected in 31.8%. AG ≥16.8% presented similar accuracy for detecting DM as defined by a HbA1c >6.5%. The use of the 2-h plasma glucose after a 75-g OGTT and GA (2hPG/GA) ratio improves sensitivity, reduces the number of incorrect diagnoses by GA or HbA1c >6.5% and has an accuracy comparable to OGTT, indicating that the use of approach applying the postprandial glucose (PPG) and GA (PPG/GA) may be more convenient for patients and increase the diagnostic performance of the test. Studies to validate this approach are needed.

A prevalência de diabetes mellitus (DM) está aumentando constantemente em todo o mundo a uma taxa alarmante. Devido às suas complicações que causam uma maior morbidade, invalidez e mortalidade, o DM representa um enorme problema para a saúde pública. Com isso, ações, tanto em diagnóstico como em tratamento, são necessárias para desacelerar a tendência atual e prevenir o desenvolvimento das complicações do DM. Recentemente, a hemoglobina glicada (HbA1c) foi introduzida nos critérios diagnósticos de DM. Os resultados de HbA1c são igualmente apropriados para o diagnóstico, apesar de não necessariamente detectarem DM nos mesmos indivíduos detectados pelos critérios de glicemia. No entanto, os níveis de HbA1c podem ser influenciados por qualquer condição que altere a vida útil dos eritrócitos e metabolismo da hemoglobina, independentemente da glicemia, resultando em erro diagnóstico neste grupo da população com estas condições. Além disso, estudos epidemiológicos revelaram que a glicemia pós-prandial tem um maior risco de causar complicações cardiovasculares em relação à hiperglicemia persistente e pode ser acessada com precisão usando a albumina glicada (AG) e não a HbA1c. Nesse contexto, estudos recentes têm evidenciado que a AG pode ser um marcador para diagnóstico do DM e também ser utilizado como um marcador alternativo à HbA1c para o controle glicêmico. No entanto, esses estudos, foram realizados apenas na população asiática e podem não ser aplicáveis a outros grupos étnicos. Por isso, mais investigações para a validação do desempenho diagnóstico da AG na predição do DM em diferentes grupos etnicos são necessárias. Neste estudo avaliamos o desempenho da AG no diagnóstico do DM em 242 indivíduos brasileiros com idade média de 54,4 anos (+ 13,0). Baseando-se nos valores de glicose plasmática durante o teste oral de tolerância à glicose (TOTG), o DM foi detectado em 31,8%. AG ≥16,8% apresentou acurácia similar para a detecção de DM conforme definido por HbA1c >6,5%. O uso da razão glicemia de 2h pós-sobrecarga de 75g de glicose (2hPG) e AG (2hPG/AG) melhora a sensibilidade, reduz o número de diagnósticos incorretos por AG ou HbA1c >6,5% e possui um acurácia comparável ao TOTG, indicando que o uso de uma estratégia aplicando a razão da glicemia pós-prandial (GPP) real e AG (GPP/AG) pode ser mais conveniente para pacientes e aumentar o desempenho diagnóstico do teste. Estudos para validar esta estratégia são necessários. / The prevalence of diabetes mellitus (DM) is constantly increasing worldwide at an alarming rate. Due to its complications that cause greater morbidity, disability and mortality, DM represents a heavy burden on public health. Therefore, actions in both, diagnosis and treatment, are necessary to slow down the current tendency and avoid the development of DM complications. Recently, the glycated hemoglobin test (HbA1c) was introduced in the diagnostic criteria for DM. The HbA1c results are equally appropriate for diagnostic testing, though not necessarily detect DM in the same subjects detected by plasma glucose criteria. However, blood HbA1c levels may be influenced by any condition that changes the lifespan of erythrocytes and hemoglobin metabolism regardless of glycemia, resulting in the misdiagnosis of this population group. In addition, epidemiological studies have shown that postprandial glycemia has a higher risk of causing cardiovascular complications than chronic hyperglycemia and can be accurately assessed using the glycated albumin (GA) test rather than HbA1c. In this context, recent studies have shown that GA may be a marker for the diagnosis of DM and also be used as an alternative marker to HbA1c on many occasions. However, these studies have been conducted only in the Asian population and may not be applicable to other ethnic groups. Therefore, further investigations to validate the diagnostic performance of GA in the prediction of DM in different ethnic groups are necessary. In this study, we evaluated the GA performance in the diagnosis of DM in 242 Brazilian individuals with a mean age of 54.4 years (+ 13.0). Based on plasma glucose values during oral glucose tolerance test (OGTT), DM was detected in 31.8%. AG ≥16.8% presented similar accuracy for detecting DM as defined by a HbA1c >6.5%. The use of the 2-h plasma glucose after a 75-g OGTT and GA (2hPG/GA) ratio improves sensitivity, reduces the number of incorrect diagnoses by GA or HbA1c >6.5% and has an accuracy comparable to OGTT, indicating that the use of approach applying the postprandial glucose (PPG) and GA (PPG/GA) may be more convenient for patients and increase the diagnostic performance of the test. Studies to validate this approach are needed.