Question 4Are the symptoms and description of the symptoms different in women presenting with acute chest pain of suspected cardiac origin compared with men

Grading: 1-Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias*
Shaw LJ;Bairey Merz CN;Pepine CJ;Reis SE;Bittner V;Kelsey SF;Olson M;Johnson BD;Mankad S;Sharaf BL;Rogers WJ;Wessel TR;Arant CB;Pohost GM;Lerman A;Quyyumi AA;Sopko G;
Insights from the NHLBI-Sponsored Women's Ischemia Syndrome Evaluation (WISE) Study: Part I: gender differences in traditional and novel risk factors, symptom evaluation, and gender-optimized diagnostic strategies
Ref ID 10303J Am Coll Cardiolpgs: S4 to S202006
Study TypeSystematic ReviewFundingNational Heart, Lung and Blood institute; National Centre for Research Resources; Gustavus and Louis Pfeiffer Research Foundation; Womens Gulid of Cedars-Sinai Medical Centre; Ladies Hospital Aid Society of Western Pennsylvania
Number of participants195 Studies,
Inclusion/Exclusion Criteria
Patient Characteristics
Recruitment
Setting
Interventions/Test/Factor being investigated
Comparisons
Length of Study/Follow-up
Outcome measures studied
Results
Safety and adverse effects
Does the study answer the question?The study reviews papers on the presenting symptoms and links to MI and obstructive coronary disease. The review suggests, despite there being differences in the symptoms women present with; symptoms evaluation in women had not been fully evaluated due to studies often applying typical angina definitions which were defined through male populations to females. These differences are seen in the frequency, type and quality of symptoms. The study reviews evidence which shows that initial symptoms in women often include fatigue, sleep disturbance, and shortness of breath.
The review states that a recent study reported no differences in the accuracy of typical symptoms, defined as chest pain or discomfort, dyspnea, diaphoresis, and arm or shoulder pain between men and women when diagnosing ACS. However chest pain/discomfort and diaphoresis were the most commonly presented symptoms in women who had a confirmed diagnosis of ACS. Women were also more likely to report acute initial symptoms but up to half of the women had no prior chest pain symptoms when diagnosed with AMI. The review reports that women are less likely to present with exertional chest pain (typical angina) than men but were more likely to be admitted to hospital for chest pain than men (4 million visits for women vs. 2.4 million for men). The review suggests from this evidence that when assessing chest pain in women the effect exertion has on symptoms should be taken into account for defining typical angina. The review states that the Yale group's definition of angina (which includes chest pain or discomfort, dyspnea, diaphoresis, and arm or shoulder pain) gives an accurate method of identifying unstable angina, however other studies have included exertional components to the symptoms which leads to more accurate diagnosis.

The review states that older women are more likely to present with symptoms similar to men compared with younger women, which could be explained by the fact that older women have more typical angina. There are no differences in the rate of diagnosis of ACS in older men and women; however women aged under 65 are more likely to be discharged without a diagnosis of unstable angina, who are also less likely to have ST-segment elevation MI, the review suggests that this can protract their time to diagnosis and the intensity of management and can lead to poorer outcomes.

The review went on to analyse the presenting symptoms which are suggestive of MI, women were less likely to have obstructive CAD than men on angiography, which was first highlighted by Diamond and Forrester in the 1980's. This study showed that women with typical and atypical chest pain symptoms have been used to calculate the probably of a women having obstructive CAD being considerably less than that for a man. The review gives the example of “typical exertional angina in a 55 year old man has a probability of obstructive CAD of approximately 90% as compared with a wide range from 55-90% for a 55 year old woman”. The review reports that this leads the conclusion that the use of chest pain symptoms to diagnose obstructive CAD in a woman is not as accurate as for a man. This conclusion and trend of symptoms being inaccurate at diagnosing obstructive CAD by Diamond and Forrester has been reported in later studies with other female populations, especially in women with a history of diabetes. The review states that this could be due to the descriptors of symptoms used by women, as those who report stable or intermittent chest pain, the description of the chest pain is a doctor's most important diagnostic tool which may lead to less intensive management.

The review highlights 2 questions to be answered which current evidence is unable to do: “can current symptom evaluation tolls be improved for more accurate detection of obstructive CAD in women? Do symptom differences suggest s gender-specific pathophysiology such that gender-specific new tools should be developed for the assessment of IHD in women?”

The review stated that the most women who had a coronary angiography which did not show obstructive CAD continued to have symptoms which lead to a poor quality of life and who continued to require repeated health investigations. The study reported that this required many doctors to use cardiac imaging to differentiate cardiac and noncardiac symptoms. The review concludes that this method does not give a technique to identify and manage myocardial ischemia in women who do not have significant obstructive CAD.

The review continued to assess postmenopausal women to show that they are likely to have a cluster of risk factors including hypertension, obesity and dyslipidemia. The study suggests this could be related to gender-specific differences in metabolic rate which is increased due the hormonal imbalances caused by the menopause. This shows a cluster of risk conditions which include insulin resistance (with or without glucose intolerance), dyslipidemia (elevated triglycerides, small LDL particles, or low HDL cholesterol), hypertension, and obesity. The study refers to the National Cholesterol Education Program Adult Treatment Panel-III which has a simplified the definition of clustering risk factors to the presence of 3 or more of “1) waist circumference >35 inches; 2) fasting triglycerides >150 mg/dl; 3) HDL cholesterol <50 mg/dl; 4) hypertension (systolic blood pressure ≥130mmHg, diastolic blood pressure ≥85 mmHg, or use of antihypertensive drug therapy); or 5) a fasting glucose measurement ≥110mg/dl”. The authors state the evidence has shown that obesity is not an independent predictor of cardiovascular disease but the metabolic syndrome leads to a link between cardiovascular disease and obesity.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?
Internal Validity
Grading: 2++High-quality systematic reviews of case–control or cohort studies High-quality case–control or cohort studies with a very low risk of confounding, bias or chance and a high probability that the relationship is causal
Pryor DB;Harrell FE;Lee KL;Califf RM;Rosati RA;
Estimating the likelihood of significant coronary artery disease
Ref ID 10283The American journal of medicinepgs: 771 to 7801983
Study TypeCohortFundingNot reported
Number of participants3627 in training population, 1811 in test population
Inclusion/Exclusion CriteriaPatients had progressive chest pain in the frequency, severity or duration had increased in the 6 weeks prior to catherisation or preinfarctional chest oan which had a very unstable pain pattern that resulted in admission to the coronary care unit for evaluation of the possible MI
Patient CharacteristicsPatient characteristics which were collected were:
History: age, sex, chest pain history (pain type, severity, frequency, nocturnal, pregressive, preinfarctional), duration of CAD, preevious history of MI, congestive heart failure, history of vascular disease
Risk factors: smoking, hyperlipidaemia, hypertension, diabetes, family history
Physical examination: ventricular gallop, systolic blood pressure
ECG: ST-T wave changes, electrocardiographic premature ventricular contractions, Electrocardiographic Q waves
Chest X-Ray: cardiomegaly
RecruitmentPatients admitted for cardiac catherisation between 1969 and 1982.
SettingSecondary care, USA
Interventions/Test/Factor being investigatedDiagnosis of chest pain.
ComparisonsPatient characteristics which give a probability of disease
Length of Study/Follow-up
Outcome measures studiedProbability of disease
ResultsThe study had a training population of 3627 patients who were seen between 1969 and January 1979, from these patients a stepwise logistic regression analysis was used to develop a model for predicting the probability of significant CAD. A test population of 1811 patients seen between January 1969 and January 1982, in this population the model developed in the test population was used to predict the probability of CAD for each patient.
The authors then tested the model in other populations (from CASS study) to estimate the prevalence of disease in subgroups of the patients in the literature (external validation)

Results from training population:
Clinically Important Characteristics and the Chi-squared:
Pain type (typical, atypical or nonanginal) – 1091
Previous MI – 511
Sex – 187
Age – 119
Smoking – 79
Hyperlipidaemia – 26
ST-T wave changes – 28
Diabetes – 12
Interactions
age × sex
age × smoking
age × hyperlipidaemia
sex × smoking

Poor Clinical Predictors of Significant CAD and the Chi-squared:
Chest pain severity – 0.96
Chest pain frequency – 8.57
Nocturnal chest pain – 2.22
Progressive chest pain – 2.54
Preinfarction angina – 9.70
Vascular disease – 0.40
Duration of CAD – 9.16
Congestive heart failure – 0.59
Hypertension – 5.19
Family history – 6.39
Ventricular gallop – 1.06
Cardiomegaly – 1.41
Electrocardiographic premature ventricular contractions – 0.46

The results from the training group are shown under “Clinically Important Characteristics and the Chi-squared” in the order of their importance (chi-squared added to the model by the parameter, adjusting for the characteristics that precede it). The type of chest pain (typical, atypical or nonanginal) was the most important characteristic followed by previous MI, sex, age, smoking, hyperlipidaemia, ST-T wave changes on ECG, diabetes. The table shows the 4 significant interactions which were found.
The study also showed that in men the effect of an increasing age was more important than in women, smoking was more important for women than men, and that smoking and hyperlipidaemia were more important at younger ages. The results for the other characteristics which were found to have small or nonsignificat effects on the prevalence of disease are shown under “Poor Clinical Predictors of Significant CAD and the Chi-squared”

The authors then validated the model in the test population which showed that the predicted probability of disease is nearly identical to the observed prevalence. This was with the exception of the group with predicted estimates of 0.475 to 0.525 (this group 8 out of 34 patients, with significant disease). The median prediction for patients with disease was 94% compared with a median prediction of 33% for patients without disease. A predicted probability of significant disease > 0.83 was found in 75% of patients with disease and in less than 10% of patients with disease. A probability of significant disease < 0.33 was found in nearly 50% of patients without disease and in less than 5% of patients with disease.

The authors then externally validated using the population from the CASS study. There was disagreement on patients classified as having nonanginal chest pain (where the greatest difference in predicted disease compared to observed disease was seen), but the predicted estimates from the model were nearly equal to the observed prevalence of disease. The predicted estimates from the model of the probability of significant disease were nearly identical to the observed prevalence for subgroups based on “age, sex and history of MI” or “age, sex and pain type”.
Safety and adverse effectsNone
Does the study answer the question?The results from the training population showed the type of chest pain (typical, atypical or nonanginal) was the most important characteristic followed by previous MI, sex, age, smoking, hyperlipidaemia, ST-T wave changes on ECG, diabetes. The study also showed that in men the effect of an increasing age was more important than in women, smoking was more important for women than men, and that smoking and hyperlipidaemia were more important at younger ages. The study also found some characteristics to have small or nonsignificat effects on the prevalence of disease.

The authors then validated the model in the test population which showed that the predicted probability of disease is nearly identical to the observed prevalence. When comparing the model to an external population the study showed that the predicted estimates from the model were nearly equal to the observed prevalence of disease.

The predicted estimates from the model of the probability of significant disease were nearly identical to the observed prevalence for subgroups based on “age, sex and history of MI” or “age, sex and pain type”. However the greatest difference in predicted disease compared to observed disease was seen in patients with nonanginal chest pain.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had chest pain, directly applicable to guideline.
Internal ValidityWell covered
Grading: 2+Well-conducted case–control or cohort studies with a low risk of confounding, bias or chance and a moderate probability that the relationship is causal
Griffiths DH;Pokorny ME;Bowman JM;
Differences in African American and white women with myocardial infarction: history, presentation, diagnostic methods, and infarction type
Ref ID 1293American journal of critical care : an official publication American Association of Criticalpgs: 101 to 1041999
Study TypeCohortFundingNot reported
Number of participants46, of which 18 were african-american, 28 were white
Inclusion/Exclusion Criteriawomen diagnosed with MI between January and June 1995
Patient CharacteristicsThe average age for african-american women was 66.6±14.3 years and for white women 69.1±14.2 years, the age range for all patients was 39-94 years.
RecruitmentPatients who presented with chest pain to a tertiary care facility in North Carolina, USA
Settingtertiary care facility in North Carolina, USA
Interventions/Test/Factor being investigateddifferences in african-american and white women with MI
Comparisonsdifferences in african-american and white women with MI
Length of Study/Follow-upNot reported
Outcome measures studiedRisk factors and ECG changes
ResultsPatients were initally diagnosed with a 12-lead ECG, if the initial ECG was non-diagnositc other methods included subsequent ECG, echocardiography, coronary angiography, measurement of serum levels of cardiac enzymes and other methods.

Admitting diagnosis of:
MI – 33% African American, 36% White, 35% total
Rule out MI – 11% African American, 32% White, 24% total
Angina – 17% African American, 11% White, 13% total
Other 39% African American, 21% White, 28% total

Types of MI and diagnostic methods:
Initial 12-lead ECG – Q wave 6 African American, 13 White, non-Q wave 12 African American, 15 White
Subsequent ECG – Q wave 1 African American, 1 White, non-Q wave 0 African American, 2 White
Echocardiography – Q wave 1 African American, 1 White, non-Q wave 0 African American, 0 White
Coronary angiography – Q wave 0 African American, 0 White, non-Q wave 1 African American, 0 White
Measurement of cardiac enzyme levels – Q wave 1 African American, 1 White, non-Q wave 10 African American, 11 White
Other – Q wave 0 African American, 1 White (sudden ventricular fibrillation), non-Q wave 0 African American, 1 White (history and physical examination)

Medical history variables:
Previous MI – 28% African American, 29% White, (P=1.000)
Angina – 11% African American, 29% White, (P=0.300)
Congestive heart failure – 28% African American, 29% White, (P=1.000)
Percutaneous transluminal coronary angioplasty – 11% African American, 0% White, (P=0.287)
Coronary artery bypass graft – 11% African American, 7% White, (P=1.000)
Stroke – 28% African American, 4% White, (P=0.027)
Diabetes – 56% African American, 29% White, (P=0.128)
Hypertension – 100% African American, 54% White, (P=0.002)
Current smoker – 17% African American, 21% White, (P=0.986)
Family history of coronary artery disease – 17% African American, 29% White, (P=0.568)
Hypercholesterolemia – 28% African American,18% White, (P=0.667)
Safety and adverse effectsNone
Does the study answer the question?24 patients presented with chest pain (52%), 9 of the 18 African American women (50%) and 15 of 28 white women (54%), this difference was not significant. The results for the diagnosis on admission to hospital were MI in 16 patients, rule out MI in 11 patients, angina in 6 patients and other 13 patients. The other diagnosis included 1 patients with congestive heart failure 1 with a hip fracture, 1 with decreased level of consciousness and 10 with unspecified n=10. There were no significant differences were found between African American and white women in the diagnosis on admission.

In the whole sample population those with a history of MI were more likely to have a non-Q wave than Q wave MI (n=13). In white women those with a history of MI or a history of congestive heart failure had a higher occurrence of non-Q wave then Q wave MI (both n=8). In African American women those with a history of angina had a higher occurrence of Q wave than non-Q wave MI (n=2).

At the time of admission 2 of the medical history variables were shown to be significantly different: stroke (P=0.027) and hypertension (P=0.002).
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?52% presented with chest pain. On admission, 16 patients had AMI, 11 to rule out AMI, 6 angina, 1 congestive heart failure, 1 hip fracture, 1 decreased level of consciousness, 10 other diagnosis
Internal ValidityWell covered
McSweeney JC;Cody M;Sullivan P;Elberson K;Moser DK;Garvin BJ;
Women's early warning symptoms of acute myocardial infarction
Ref ID 10299Circulationpgs: 2619 to 26232003
Study TypeCohortFundingNational Institute of Nursing Research
Number of participants515 women
Inclusion/Exclusion CriteriaWomen who were diagnosed with AMI and discharged in the previous 4-6 months from 5 sites in Arkansas, North Carolina and Ohio, Patients needed to be cognitively intact, speak english, and have telephone access
Patient CharacteristicsThe study included 515 women with an average age of 66.4±12 years. Of the 515 women 93% were white, 6.2% black, 2% Native American. For 72% of the women had no prior history of MI, the other 28% gave details of their most recent AMI.
RecruitmentPatients were those diagnosed with AMI and discharged in the previous 4-6 months from 5 sites in Arkansas, North Carolina and Ohio
SettingSecondary care, USA
Interventions/Test/Factor being investigatedsymptoms and risk factors for those with AMI
Comparisonssymptoms and risk factors
Length of Study/Follow-upNot reported
Outcome measures studiedsymptoms and risk factors
ResultsSee table 1 and 2 in McSweeney, 2003 doccument
Safety and adverse effectsNone
Does the study answer the question?The study included 515 women with an average age of 66.4±12 years. Of the 515 women 93% were white, 6.2% black, 2% Native American. For 72% of the women had no prior history of MI, the other 28% gave details of their most recent AMI.

The study considered both initial (prodromal) symptoms and acute symptoms. The average number of initial symptoms experienced was 5.71±4.36, with the most common being unusual fatigue, sleep disturbance, shortness of breath, indigestion, and anxiety. 44% of those reporting sleep disturbances and 42% of those reporting fatigue described them as severe. 29.7% of women reported chest pain/discomfort (aching, tightness, pressure, burning, sharpness fullness or tingling), with the location and descriptors used not being mutually exclusive. 78% of women reported having had at least one of their initial symptoms daily or several times a week for more than 1 month.
The average number of acute symptoms experienced was 7.3±4.8, with the most common being shortness of breath, weakness, unusual fatigue, cold sweat, and dizziness. The women reported discomfort in their back and high chest as the most common locations of pain. Again chest pain/discomfort was reported by women (pressure, ache, and tightness), mostly being described as severe pain/discomfort. Over all 43% of women reported no chest pain/discomfort.

The study also considered the risk factors; most women had a family history of cardiovascular disease, a history of cardiovascular disease and had diabetes. The average BMI was 28.6±6.5 and less than half of the women did regular exercise before having their AMI.

The study carried out multiple regression analysis to assess if the acute score could be predicted from the prodromal score. “The prodromal score accounted for an additional 33.2% of the variance in acute symptom scores after control for risk factors which accounted for only 9.9% of the variance”.

The study also carried out a T test to determine the association of symptoms with risk factors. The T test showed that there was significant association between initial symptoms and all risk factors except age >50 years, hypertension and hyperlipidemia. The T test also showed that there was significant association between acute symptoms and all risk factors except hypertension, hyperlipidemia and second hand smoke.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had AMI
Internal ValidityWell covered
Meischke H;Larsen MP;Eisenberg MS;
Gender differences in reported symptoms for acute myocardial infarction: impact on prehospital delay time interval
Ref ID 5613Am J Emerg Medpgs: 363 to 3661998
Study TypeCohortFundingNot reported
Number of participants4,497, 2970 men and 1527 women
Inclusion/Exclusion CriteriaPatients with a confirmed MI, admitted between January 1991 and February 1993 to the coronary care units of 16 King County hospitals. Those who had cardiac arrest, coma, and shock were excluded
Patient CharacteristicsGender – 66% men, 34% women
Median age – 64 years men, 73 years women (P=<0.001)
White – 91% men, 93% women
Black – 4% men, 4% women
Asian/Pacific Islander – 5% men, 3% women
RecruitmentConsecutive patients admitted between January 1991 and February 1993 to the coronary care units of 16 King County hospitals with AMI were assessed for inclusion.
SettingSecondary Care, USA
Interventions/Test/Factor being investigatedrisk factors and medical history of men and women with AMI
Comparisonsrisk factors and medical history of men and women with AMI
Length of Study/Follow-upNot reported
Outcome measures studiedrisk factors (gender, age, race, history of AMI, history of diabetes) medical history (chest pain symptoms, diaphoresis, dyspnea, epigastic pain, nausea/vomiting, syncope)
ResultsUnivariate comparison of medical history and symptoms:
Gender – 66% men, 34% women
Median age – 64 years men, 73 years women (P=<0.001)
White – 91% men, 93% women
Black – 4% men, 4% women
Asian/Pacific Islander – 5% men, 3% women
History of AMI – 30% men, 26% women (P=0.021)
History of diabetes – 19% men, 25% women (P=<0.001)
Chest pain symptoms – 92% men, 89% women (P=<0.001)
Diaphoresis – 54% men, 44% women (P=<0.001)
Dyspnea – 46% men, 52% women (P=<0.001)
Epigastric pain – 11% men, 11% women, Not significant
Nausea/vomiting – 35% men, 44% women (P=<0.001)
Syncope – 3% men, 3% women, Not significant

Beta and P value regression for medical history and symptoms:
Age – β 0.096, P=<0.001
Gender – β 0.053, P=0.002
History of AMI – β -0.064, P=<0.001
History of diabetes – β 0.048, P=0.004
Diaphoresis – β -0.147, P=<0.001
Chest pain – β -0.059, P=<0.001
Syncope – β -0.039, P=0.02
Dyspnea – β -0.024, Not significant
Epigastric pain – β 0.03, Not significant
Nausea/vomiting – β 0.014, Not significant
Safety and adverse effectsNone
Does the study answer the question?This study showed that women were significantly older than men and were more likely to have a history of diabetes. Women were also more likely to report sweating and nausea, this difference persisted after adjustment for age and history of diabetes. Women were also more likely to report shortness of breath, especially younger women and those who had a history of diabetes.
Men were more likely to have a history of AMI than women. There was no difference between men and women in presentation of chest pain, this similarity persisted after adjustment for age and history of diabetes.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had a confirmed AMI
Internal ValidityWell covered
Milner KA;Funk M;Arnold A;Vaccarino V;
Typical symptoms are predictive of acute coronary syndromes in women
Ref ID 10301Am Heart Jpgs: 283 to 2882002
Study TypeCohortFundingPart funded by Ethel F. Donoghue Women's Health Investigation Program at Yale
Number of participants522 in total, 246 women and 276 men
Inclusion/Exclusion Criteriaaged 45 years or older, reportd at least one prespecified set of typical or a typical symptoms suggestive of ACS
Patient CharacteristicsThe mean age for women with ACS was 69 ± 15 years, the mean age for women without ACS was 64 ± 15 years,
RecruitmentPatietns who were seen in the emergency department with suspected ACS
SettingSecondary Care, USA
Interventions/Test/Factor being investigatedrisk factors and symptoms of women and men presenting with suspected ACS
Comparisonsrisk factors and symptoms of women and men presenting with suspected ACS
Length of Study/Follow-upNot reported
Outcome measures studiedRisk factors and clinical history of patients
ResultsBaseline characteristics:
White race – 36% women with ACS, 46% men with ACS
History of coronary heart disease – 44% women with ACS, 48% men with ACS
Systemic hypertension – 38% women with ACS, 49% men with ACS
Obesity – 38% women with ACS, 46% men with ACS
History of MI – 49% women with ACS, 51% men with ACS
Diabetes – 47% women with ACS, 46% men with ACS
Hypercholesterolemia – 41% women with ACS, 50% men with ACS
Other cardiac problems – 39% women with ACS, 35% men with ACS
History of heart failure – 40% women with ACS, 45% men with ACS
Current smoker – 26% women with ACS, 42% men with ACS

Relationship between typical symptoms and ACS:
Chest pain/discomfort present in – 36% women with ACS, 49% men with ACS
Dyspnea present in – 44% women with ACS, 41% with ACS
Arm or shoulder pain present in – 38% women with ACS, 47% with ACS
Diaphoresis present in – 53% women with ACS, 44% with ACS
Neck or jaw pain present in – 41% women with ACS, 53% with ACS

Relationship between atypical symptoms and ACS:
Nausea or vomiting present in – 39% women with ACS, 48% men with ACS
Dizziness present in – 36% women with ACS, 32% men with ACS
Indigestion present in – 38% women with ACS, 45% men with ACS
Fatigue present in – 36% women with ACS, 41% men with ACS
Chest fullness, stabbing, numbness, burning or right chest pain present in – 34% women with ACS, 50% men with ACS
Midback pain present in – 50% women with ACS, 17% men with ACS
Palpitations present in – 35% women with ACS, 29% men with ACS
Upper-extremity numbness present in – 29% women with ACS, 33% men with ACS
Unable to take a deep breath present in – 9% women with ACS, 29% men with ACS
Cough present in – 25% women with ACS, 40% men with ACS

Symptom predictors of ACS in women and men by logistic regression analysis:
(relative risk – RR)
Women
Chest pain or discomfort – RR – 1.81, 95% CI 0.95 to 3.42, P=0.069
Neck or jaw pain – RR – 1.60, 95% CI 0.83 to 3.10, P=0.163
Diaphoresis – RR – 2.53, 95% CI 1.17 to 5.48, P=0.019
Men
Chest pain or discomfort – RR – 1.56, 95% CI 0.86 to 2.82, P=0.142
Neck or jaw pain – RR – 0.69, 95% CI 0.40 to 1.19, P=0.182
Diaphoresis – RR – 0.49, 95% CI 0.26 to 0.93, P=0.028

Relative risk of ACS for typical symptoms in women relative to men:
Chest pain or discomfort – RR – 0.83, 95% CI 0.66 to 1.06, P=0.129
Neck or jaw pain – RR – 0.69, 95% CI 0.40 to 1.15, P=0.141
Diaphoresis – RR – 1.18, 95% CI 0.87 to 1.59, P=0.384
Arm or shoulder pain – RR – 0.91, 95% CI 0.64 to 1.30, P=0.612
Dyspnea – RR – 1.00, 95% CI 0.74 to 1.35, P=0.993
Safety and adverse effectsNone
Does the study answer the question?The study showed that older women and men were both significantly more likely to be diagnosed with ACS than younger men and women. Women with a history of coronary heart disease, MI or diabetes were also significantly more likely to be diagnosed with ACS compared to those without the risk factors. Men without a history other cardiac problems were more likely to be diagnosed with ACS. Women who were diagnosed with ACS had a higher number of symptoms than those without (3.36±1.74 compared to 2.78±1.46 P=0.006), however there was no difference in the number of symptoms for men with ACS compared to men without ACS. Typical symptoms in men were not significantly related to a diagnosis of ACS, however those with dizziness or fainting were less likely to be diagnosed with ACS. Women with typical symptoms (chest pain or discomfort, diaphoresis, dyspnea and arm or shoulder pain) were significantly more likely to be diagnosed with an ACS. A multivariate analysis of independent predictors of ACS showed that diaphoresis was strongest in predicting ACS in women, followed by chest pain or discomfort (81% higher risk for ACS) and arm or should pain had a (60% higher risk for ACS). The model for male patients was a poor fit, the authors suggested that this meant that a patients symptoms were not a useful predictor of ACS.

The study went on to compared men and women, which showed that there was no difference in the typical symptoms for men and women. The study showed that there were no sex differences through comparing the adjusted the relative risks for ACS in women with typical symptoms and in with men with typical symptoms which was both close to 1.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had symptoms suggestive of ACS
Internal ValidityWell covered
Penque S;Halm M;Smith M;Deutsch J;Van RM;McLaughlin L;Dzubay S;Doll N;Beahrs M;
Women and coronary disease: relationship between descriptors of signs and symptoms and diagnostic and treatment course
Ref ID 10292American journal of critical care : an official publication American Association of Criticalpgs: 175 to 1821998
Study TypeCohortFundingNot reported
Number of participants98 patients, of which 51 were women and 47 were men
Inclusion/Exclusion CriteriaIncluded: primary medical diagnosis of MI, at least 21 years old, english speaking, admission via emergency department, directly from physician's office or by transfer from rural hospital within 6 hours of MI. Exclusion: patients who had sudden cardiac death events. A history of coronary artery disease was not a reason for exclusion and so the population is mixed
Patient CharacteristicsThe mean age for all patients was 59 years. For the women the mean age was 61 years (range 41-89 years), for the men the mean age was 56 years (range 37-79 years). 3% of all patients were uninsured (measure of socio economic status)
Recruitmentadmitted to the hospital during a period of 12 months, with a primary diagnosis of MI
Settingsecondary care, USA
Interventions/Test/Factor being investigateddifferences between men and women in signs and symptoms of MI
ComparisonsMena and women
Length of Study/Follow-upNot reported
Outcome measures studiedrisk factors, signs and symptoms
ResultsCardiovascular risk factor profile
Family history of heart disease – women 56%, men 51%
Past or current history of smoking – women 57%, men 81%
Hypertension – women 41%, men 46%
Hyperlipidaemia – women 49%, men 55%
Diabetes – women 20%, men 17%

Precipitating factors for chest pain
Rest – women 53%, men 55% (P=0.89)
Exertion – women 63%, men 40% (P=0.09)
Sex – women 10%, men 6% (P=0.40)
Stress – women 51%, men 34% (P=0.10)

Time elapsed after cardiac-related signs or symptoms were first experienced before treatment was sort
Less than 24 hours – women 15%, men 22%
1-2 days – women 6%, men 9%
3-7 days – women 15%, men 17%
8-30 days – women 15%, men 15%
2-6 months – women 6%, men 13%
6-12 months – women 6%, men 0%
More than 1 year – women 38%, men 24%

Descriptors of associated signs and symptoms
Fatigue – women 71%, men 70% (P=0.90)
Rest pain – women 71%, men 72% (P=0.80)
Weakness – women 68%, men 62% (P=0.60)
Shortness of breath – women 66%, men 66% (P=0.70)
Dizziness – women 56%, men 43% (P=0.10)
Arm pain – women 53%, men 55% (P=0.70)
Nausea – women 51%, men 35% (P=0.10)
Back pain – women 52%, men 20% (P=0.005)
Loss of appetite – women 43%, men 19% (P=0.03)
Neck pain – women 41%, men 35% (P=0.10)
Sweating – women 48%, men 42% (P=0.60)
Heartburn – women 28%, men 33% (P0.50=)
Paroxysmal nocturnal dyspnea – women, 30% men 11% (P=0.05)
Palpitations – women 25%, men 26% (P=0.80)
Jaw pain – women 10%, men 13% (P=0.90)
Throat pain – women 8%, men 22% (P=0.10)
Toothache – women 5%, men 2% (P=0.40)
Safety and adverse effectsNone
Does the study answer the question?The study considered the descriptors of signs and symptoms. The study showed that chest discomfort was the most common initial symptom reported by both men (51% as an initial symptom, 99% at some point) and women (49% as an initial symptom, 94% at some point). The 4 most reported symptoms for men and women were fatigue, rest pain, weakness, and shortness of breath, however women reported dizziness and men reported arm pain as the next common symptom. Women were more likely to suffer loss of appetite, paroxysmal nocturnal dyspnea and back pain than men. These differences were significant: loss of appetite (chi-squared=4.48), paroxysmal nocturnal dyspnea (chi-squared=3.80), and back pain (chi-squared=7.60).

The study considered the length of time from initial symptoms to seeking medical help. There was no significant difference between men (5.3 hours) and women (4.2 hours), with the majority of men and women first having symptoms in the preceding 24 hours, the previous 3 days to 1 month or more than 1 year before. The study also considered the mean number of words used to describe signs, there was no significant difference between men (58) and women (55).

The study concluded that “chest pain was the first sign or symptom of MI reported by both men and women”. Women were more likely to report back pain, loss of appetite, and paroxysmal nocturnal dyspnea as symptoms than men and were less likely than men to have diagnostic angiography and to receive IV nitroglycerin, heparin, and thrombolytics as part of their management.
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had a primary diagnosis of MI
Internal ValidityWell covered
Grading: 2-Case–control or cohort studies with a high risk of confounding bias, or chance and a significant risk that the relationship is not causal*
De S;Searles G;Haddad H;
The prevalence of cardiac risk factors in women 45 years of age or younger undergoing angiography for evaluation of undiagnosed chest pain
Ref ID 923The Canadian journal of cardiologypgs: 945 to 9482002
Study TypeCohortFundingNot reported
Number of participants187 in total, 55 in group A (those with significant CAD) 132 in group B (those without significant CAD)
Inclusion/Exclusion CriteriaWomen aged under 45 years, who were referred for coronary angiography due to chest pain and who had no known history of CAD
Patient CharacteristicsNot reported.
Patients were women aged under 45 who did not have a known history of CAD
RecruitmentPatients referred for coronary angiography due to chest pain during a 4 year period (february 1997-December 2000) at Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia
SettingSecondary care, Nova Scotia, Canada
Interventions/Test/Factor being investigatedRisk factors in women with and wothout signficant CAD
ComparisonsRisk factors - obesity, dyslipidemia, diabetes, hypertension, premature family history of CAD, current smoker, past smoker
Length of Study/Follow-upNot reported
Outcome measures studiedDiagnosis of CAD
ResultsRisk factors:
Obesity – 45% group A, 46% group B, P=0.92
Dyslipidemia – 72% group A, 47% group B, P=0.002
Diabetes – 29% group A, 9% group B, P=<0.001
Hypertension – 40% group A, 28% group B, P=0.13
Family history of premature CAD – 65% group A, 67% group B, P=0.79
Current smoker – 55% group A, 35% group B, P=0.03
Past smoker – 13% group A, 15% group B, P=0.03
Safety and adverse effectsNone
Does the study answer the question?The women included were aged <45 years that were referred for coronary angiography due to chest pain but had not been diagnosed and had no history of CAD, the patients were subsequently divided into two groups; dependant upon the presence of CAD or absence. Group A had significant CAD, and group B were without significant CAD. Group B (those without significant CAD) was subdivided into those with noncritial CAD (8%) and those with normal coronary arteries (92%). Group A were significantly more likely to have dyslipidemia (72% group A, 47% group B, P=0.002), diabetes (29% group A, 9% group B, P=0.001), and to smoke (67% group A, 50% group B, P=0.03). There was no significant difference between group A and B in the rates of obesity, hypertension, and family history of premature CAD.

The study concluded that women with CAD were more likely to have dyslipidemia, diabetes and smoking. However for women with and without CAD the commonest risk factor was a family history of CAD (67%), followed by smoking (55%) and dyslipidemia (55%).
Effect due to factor in study?Yes
Consistency of results with other studies?Consistent
Directly applicable to guideline population?Patients had chest pain
Internal ValidityWell covered

From: Appendix D, Clinical evidence extractions

Cover of Chest Pain of Recent Onset
Chest Pain of Recent Onset: Assessment and Diagnosis of Recent Onset Chest Pain or Discomfort of Suspected Cardiac Origin [Internet].
NICE Clinical Guidelines, No. 95.
National Clinical Guideline Centre for Acute and Chronic Conditions (UK).
Copyright © 2010, National Clinical Guideline Centre for Acute and Chronic Conditions.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.