Journal Reading 報告者：陳輝財 醫師 指導者：張志華 醫師 90-12-15 Presentation and Symptom Predictors of Coronary Heart Disease in Patient With and Without Diabetes American Journal of Emergency Medicine, Oct 2001, 482-487 Introduction Coronary Heart Disease is the single largest killer of Americans (459,841 deaths in 1998, or 1 in every 5 deaths in the united states). Diabetes mellitus is a strong risk factor for CHD and that CHD represents the most common cause of death for patients with diabetes (2/3 people with DM die from heart or blood vessel diseases). In patients with MI, DM is a important prognostic factor for development of heart failure, recurrent MI, and long-term mortality. Introduction CHD is often silent in patients with DM (1/4 people with DM had ECG evidence of prior MI, and 1/2 of these patients had no symptoms). In the National Registry of Myocardial Infarction-2 (NRMI-2), diabetes was independently associated with atypical presentation. Thus, there should be a higher index of suspicion for CHD for patients with DM in ED. Early detection of CHD in patient with DM can improve outcome and reduce mortality. 2 Research questions Is there a difference in symptom presentation of CHD between patients with and without diabetes ? Is there a difference in symptom predictors of CHD between patients with and without diabetes ? 3 Hypotheses Patients with diabetes will be more likely to present with atypical symptoms compared with patients without diabetes. Typical symptoms will be predictive of CHD in both patients with diabetes and patients without diabetes Atypical symptoms will be predictive of CHD in patients with diabetes, but not in patients without diabetes. Methods – Setting and Sample A prospective study in a 900-bed cardiac referral center in northeastern United states. Including Criteria Age 45 years: At least 1 symptoms suggestive of CHD. DM or least 2 cardiac risk factors Age between 18 to 44 years: Hx of CHD, HTN, Hypercholesteremia, Smoking, Obesity, Family Hx of premature CHD, Postmenopausal status. At least 1 symptoms suggestive of CHD. 528 cases (total 536 cases, 531 agreed, and information on diabetes status was missing on 3 patients ). Methods – Procedure Severity of symptoms: by using 0 to 10 scores. DM: by patients, family, or medical record Only coded as either present or absent, no information on the type of DM or DM control. CHD: acute ischemia or acute MI Acute ischemia ECG: ST depression or elevation, T inversion or T abnormalities in at least 2 consecutive leads Lack of cardiac enzyme elevation ECG change Elevated cardiac enzyme (CK-MB >5% of total CK) Acute MI Methods – Statistical Analyses Using SAS Version 8 statistical software. Symptoms reported by >5% of the sample were included in the analyses. Methods – Statistical Analyses Typical symptoms Classic chest symptoms (pain, pressure, heaviness, tightness, or squeezing in the center or left chest), neck or throat pain, jaw pain, arm or shoulder pain, diaphoresis, or SOB. Atypical chest pain (fullness, stabbing, numbness, or burning in the chest or pain in the right chest), epigastric discomfort, nausea or vomiting, upper extremity numbness or tingling, midback pain, pain with deep breath, coudh, dizziness or syncope, fatigue or weakness, and palpitatoins. Atypical symptoms Methods – Statistical Analyses For research question 1 Chi-square analysis to examine differences in symptom presentation of CHD between patients with and without diabetes ( n = 125). T-test to compare the mean rating of severity of symptoms between patients with and without diabetes. Multiple regression (with an alpha level of 0.20 for entering and staying in the model) to control for factors that could have an impact on perceived severity of symptoms (history of heart failure, age, gender, prior daignosis of CHD, psychiatric history, and smoking). Methods – Statistical Analyses For research question 2 2 subgroups: DM (n=142), no-DM (n=386). Chi-square analysis to determine symptoms associated with CHD. T-test analyses to compare the mean rating of severity of symptoms Stepwise logistic regression analysis to determine independent symptom predictors of CHD. Entered all typical and atypical symptoms reported by >5% of the sample and forced in age and gender as control factors. Alpha-level of 0.20 for entering and staying in the model and confirmed our results by forward selection and backward elimination modeling. Statistics - Chi-square analysis 卡方檢定 最常用來檢定兩組類別資料間之相關性的統計方法。（例： 比較有無DM和發生CHD此兩群組間之相關性）。 其檢定方法，即為將研究結果所計算的卡方值，和卡方分 佈下的卡方值來比較。 發生CAD 有 無 總和 2 (卡方值) = [(觀察值-預期值)2/預期值] 有無DM a的預期值 = 有 (a+b)/(c+d) 無 12 a 6 c 88 b 100 94 d 100 總和 18 182 200 若以P是否小於0.05來判定是否有統計上之意義，則看計算 出來之卡方值，若大於3.84，即有統計學上之意義。 Statistics - t-test One sample t-test, two sample t-test (independent t-test), paired t-test 檢定平均數最常用之方法 (sample 3時，則使用 ANOVA變異數分析) One sample t-test：檢定一組樣本的平均數是否和標 準值不同時使用之。即檢定得到之結果可否代表母群體 (有統計學上之意義)，或是為一隨機產生的結果。 Two sample t-test：比較兩組資料之平均數時使用之。 (例：比較有無DM之病人得到CHD的比率差距，是否有 統計學上的意義)。 Paired t-test：比較配對資料的平均數時使用之 (例： 比較同一群病人在治療前後身體狀況的差別，是否具有 統計學上的意義)。 Results Total sample of 528 patients 215: CHD 142: DM 66: DM + CHD Results - Symptom Presentation of CHD in Patients With and Without Diabetes 215 patients with CHD 80 with acute MI and 135 with acute ischemia 66 with diabets and 149 without diabetes. A majority of patients with CHD were white (73.5%) and men (58.6%). Age from 30 to 96 years (mean 65.9 14.8 years) There were no statistically significant differenced in symptom presentation between patients with and without diabetes. Results - Symptom Presentation of CHD in Patients With and Without Diabetes Using 0 to 10 scores, patients with DM (mean = 7.51 2.20) rated their symptoms as significantly more severe than those without DM (mean = 6.68 2.58, P = 0.042). The relationship between diabetes and severity of symptoms remained significant (P = 0.036) even when using multiple regression to control for factors that could have an impact on perceived severity of symptoms. Only 173 of the 215 patients with CHD (80.5%) were able to quantify the severity of their symptoms older patients were less likely to be able to quantify the severity of their symptoms (mean age = 73.4 15.2 years). The mean age of those who did rate their symptoms was 64.1 14.2(P = 0.0002). Results - Symptom Predictors of CHD in Patients With and Without Diabetes DM (n=142) A majority of patients with CHD were white (70.4%) and men (48.6%). Age from 32 to 93 years (mean 65.3 13.7 years) 66 patients with CHD and 76 patients without CHD. A majority of patients with CHD were white (78.3%) and men (54.7%). Age from 22 to 96 years (mean 62.5 16.3 years) 149 patients with CHD and 237 patients without CHD. Non-DM (n=386). Results - Symptom Predictors of CHD in Patients With and Without Diabetes Results - Symptom Predictors of CHD in Patients With and Without Diabetes Statistics - Odds ratio 定義 (勝算, odds)：一件事情發生的機率，除以這件事 情不發生的機率，定義為勝算。 Odds = p/(1-p)。 當勝算大於1時，這件事發生的機率比不發生的機率為大。 發生UTI 有無Foley Odds ratio (勝算比)： 實驗組暴露於危險因子 下之勝算，和對照組暴 露於危險因子下之勝算 的比值 。 有 15 a 5 c 20 無 總和 有 無 總和 22 b 37 18 d 23 40 60 用來估計危險因子的 相關危險度。 Odds ratio = (15x18)/(5x22) = 2.45 Results - Symptom Predictors of CHD in Patients With and Without Diabetes The perceived severity of symptoms did not differ between those with and without CHD in 2 subgroups (P = 0.504 in patient with DM and P = 0.559 in patients without DM). Only 113/142 patients with DM (79.6%) and 303/386 patients without DM (78.5%) were able to quantify the severity of their symptoms using the 0 to 10 scale. Patients with classic chest symptoms are twice as likely to be diagnosed with CHD, compared with patients without these symptoms (adjusted odds ratio = 2.10; 95% confidence interval = 1.31 - 3.37). Diaphoresis was also a significant predictor of CHD (adjusted odds ratio = 1.68; 95% confidence interval = 1.00 - 2.81; P = 0.049). In patient without DM Discussion The result of this study suggest that S/S of CHD are similar for patient with or without DM: more severe S/S in patients with DM. The severity of symptoms was not predictive of CHD in patients with or without DM. The data do not support the previous hypotheses in this study. Discussion - 3 Hypotheses Patients with diabetes will be more likely to present with atypical symptoms compared with patients without diabetes. Typical symptoms will be predictive of CHD in both patients with diabetes and patients without diabetes Atypical symptoms will be predictive of CHD in patients with diabetes, but not in patients without diabetes. Discussion - Symptom Presentation of CHD in Patients With and Without Diabetes Previous research suggests that patient with DM are more likely to have silent ischemia or silent MI. In acute MI, some studies showed patient with DM are less likely to have ischemic chest pain – maybe duo to cardiovascular autonomic neuropathy. Duo to inclusion criteria of this study (pts with S/S suggestive of CHD), there is no comment on the incidence of silent ischemia or silent MI. Discussion - Symptom Presentation of CHD in Patients With and Without Diabetes Pts in this study were equally likely to present with chest pain, regardless of diabetes status. Combine all classic chest pain together in this study – No significant difference in patient with or without DM (68.2% v 71.1%, respectively; P = 0.662). Herlitz et (7,157 patients in ED) – Slightly less often chest pain in pts with DM (90% v 93%, respectively; P <0.05) The different result may be duo to smaller sample size (n=215 with CHD) in this study and resulting lower statistical power. Discussion - Symptom Presentation of CHD in Patients With and Without Diabetes The results in this study do not indicate that Pts with DM are more likely to present with atypical symptoms. No difference in atypical chest pain (P = 0.324) No statistical significance in S.O.B (P = 0.056) More in patients with DM May be duo to a higher incidence of heart failure that found frequently in patients with DM with acute ischemia. Herlitz et Pulmonary and pleural S/S more frequently in Pts with DM – not statistical significant This study did not access other S/S Discussion - Symptom Presentation of CHD in Patients With and Without Diabetes In the present study, Pts with DM rated the severity of their symptoms higher than those without DM – contrary to the cardiovascular autonomic neuropathy hypothesis. No DM control data or physiological indicators of CV autonomic neuropathy in this study. Not all Pts with CHD were able to quantify the severity of their S/S using the 0 to 10 scores. This finding could be reflective of more extensive cardiac disease and the compounding effect of multiple system involvement in Pts with DM. Discussion - Symptom Predictors of CHD in Patients With and Without Diabetes The neck/throat pain and the arm/shoulder pain were of borderline significance in predicting CHD in Pts with DM. The classic chest symptoms (pain, pressure, heaviness, tightness, or squeezing in the center or left chest) were not predictive of CHD in Pts with DM. The presence of classic symptoms and diaphoresis were independent predictors of CHD in patients without DM. No previous investigations have examined symptom predictors of CHD in patients with DM. Discussion -Strengths and Limitations Strengths The prospective, observational design in which detailed data on presenting symptoms were collected by direct observation in the ED. Minimizes recall bias on reporting of symptoms Maximizes accuracy The potential problem of health care professionals interpreting patients’ symptoms and recording medical terms is avoided. Consider all type of typical and atypical symptoms suggestive of CHD The first study to examine symptom predictors of CHD among patients with and without diabetes. Discussion -Strengths and Limitations Limitations The relatively small number of Pts with DM diagnosed with CHD. Result in the lack of observed differenced in most of the symptoms between Pts with and without DM. No data on the type of DM, the degree of glycemic control, or indicators of CV autonomic neuropathy Some patients with DM undiagnosed – included in Pts without DM Collapse patients with acute ischemia and MI into one group duo to sample size limitations. precluding an analysis of presenting symptoms in patients with acute MI Conclusions With these strengths and limitations in mind, there were more similarities than differences in symptom presentation of CHD between patients with and without diabetes. Neck/throat pain and arm/shoulder pain: borderline significance in predicting CHD in patient with DM. Classic chest symptoms and diaphoresis: independent predictors of CHD in patients without DM. Patients with DM can present a wide variety of S/S when CHD attack. Thanks for Your Attention !!
Pages to are hidden for
"Journal Reading"Please download to view full document