Positive LRs for the diagnosis of acute myocardial infarction (AMI) are represented as means (95% CIs) of each CPC in men and women.
Pain severity in a box-and-whisker plot is quantified using a visual analog scale (VAS) in men and women with a diagnosis of acute myocardial infarction (AMI) or no AMI. Boxes represent interquartile ranges; the horizontal line in each box, the median; and whiskers, minimum and maximum values. P value indicates the comparison of the diagnostic accuracy of the VAS for AMI in women vs men.
eAppendix. Subgroup analysis of patients with first acute myocardial infarction (AMI)
eTable 1. Prevalence of the predefined chest pain characteristics in women and men with an adjudicated diagnosis of acute myocardial infarction
eTable 2. Prevalence of the predefined chest pain characteristics in women and men with an adjudicated diagnosis of ST-Segment Elevation Myocardial Infarction
eTable 3. Prevalence of the predefined chest pain characteristics in women and men with an adjudicated diagnosis of Non–ST-Segment Elevation Myocardial Infarction
eTable 4. Diagnostic performance of combinations of chest pain characteristics in women and men; univariate analysis
eTable 5. Diagnostic performance of combinations of chest pain characteristics in women and men; likelihood ratios
Rubini Gimenez M, Reiter M, Twerenbold R, Reichlin T, Wildi K, Haaf P, Wicki K, Zellweger C, Hoeller R, Moehring B, Sou SM, Mueller M, Denhaerynck K, Meller B, Stallone F, Henseler S, Bassetti S, Geigy N, Osswald S, Mueller C. Sex-Specific Chest Pain Characteristics in the Early Diagnosis of Acute Myocardial Infarction. JAMA Intern Med. 2014;174(2):241-249. doi:10.1001/jamainternmed.2013.12199
Whether sex-specific chest pain characteristics (CPCs) would allow physicians in the emergency department to differentiate women with acute myocardial infarction (AMI) from women with other causes of acute chest pain more accurately remains unknown.
To improve the management of suspected AMI in women by exploring sex-specific CPCs.
Design, Setting, and Participants
From April 21, 2006, through August 12, 2012, we enrolled 2475 consecutive patients (796 women and 1679 men) presenting with acute chest pain to 9 emergency departments in a prospective multicenter study. The final diagnosis of AMI was adjudicated by 2 independent cardiologists.
Treatment of AMI in the emergency department.
Main Outcomes and Measures
Sex-specific diagnostic performance of 34 predefined and uniformly recorded CPCs in the early diagnosis of AMI.
Acute myocardial infarction was the adjudicated final diagnosis in 143 women (18.0%) and 369 men (22.0%). Although most CPCs were reported with similar frequency in women and men, several CPCs were reported more frequently in women (P < .05). The accuracy of most CPCs in the diagnosis of AMI was low in women and men, with likelihood ratios close to 1. Thirty-one of 34 CPCs (91.2%) showed similar likelihood ratios for the diagnosis of AMI in women and men, and only 3 CPCs (8.8%) seemed to have a sex-specific diagnostic performance with P < .05 for interaction. These CPCs were related to pain duration (2-30 and >30 minutes) and dynamics (decreasing pain intensity). However, because their likelihood ratios were close to 1, the 3 CPCs did not seem clinically helpful. Similar results were obtained when examining combinations of CPCs (all interactions, P ≥ .05).
Conclusions and Relevance
Differences in the sex-specific diagnostic performance of CPCs are small and do not seem to support the use of women-specific CPCs in the early diagnosis of AMI.
clinicaltrials.gov Identifier: NCT00470587
Acute myocardial infarction (AMI) is a major cause of death and disability worldwide. Patients with symptoms suggestive of AMI account for about 10% of all emergency department (ED) consultations; even so, only 10% to 20% of them eventually receive a diagnosis of AMI. Although about 90% of patients with AMI present with chest pain and/or chest discomfort, some patients present without typical chest pain.1- 6 Rapid identification of this diagnosis is critical for the early treatment and management of AMI in these patients.2
The clinical evaluation of a patient’s medical history and symptoms is an essential cornerstone in the diagnosis of AMI, complemented by the 12-lead electrocardiogram (ECG) and results of cardiac troponin (cTn) testing.3- 6 Characterization of chest pain is an immediately available, simple, and inexpensive diagnostic tool to classify symptoms, triage patients, and initiate further diagnostic and therapeutic steps as soon as possible.7,8 Furthermore, the recent introduction of high-sensitivity cTn (hs-cTn) assays into clinical care and their associated detection of mild elevations in cTn levels for the diagnosis of AMI has reemphasized the importance of a detailed assessment of chest pain characteristics (CPCs).9- 11
Recently, sex-specific differences in symptom presentation, diagnosis, and management of cardiac disease have received increasing attention.12- 14 Most of the observed differences were to the detriment of women, including a higher rate of misdiagnosis, more delayed diagnoses, a higher AMI mortality rate, and possibly less benefit from early revascularization.15- 18 To mitigate the suboptimal outcomes of women with cardiac disease, studies have been initiated to better characterize cardiac disease in women. Some of these have revealed differences in perception and presentation of ischemic symptoms between women and men.19- 22 Although data from large AMI registries and randomized clinical trials suggested that women with AMI may differ in their symptoms from men with AMI, the important unanswered clinical question is whether detection of sex-specific CPCs is possible to allow physicians in the ED to differentiate women with AMI more accurately from women with other causes of acute chest pain.
We conducted a large, prospective, international, multicenter study of men and women presenting to the ED with acute chest pain. Our primary aim was to assess whether sex-specific CPCs would allow the differentiation of women with AMI from women with other causes of chest pain.
Advantageous Predictors of Acute Coronary Syndrome Evaluation (APACE) is an ongoing prospective international multicenter study (performed in 9 different study centers in Switzerland, Spain, and Italy) designed and coordinated by the University Hospital Basel.23- 31 The aim of the APACE study is to advance the early diagnosis of AMI. From April 21, 2006, through August 12, 2012, consecutive women and men older than 18 years presenting to the ED with symptoms suggestive of AMI with an onset or peak within the last 12 hours were recruited, after written informed consent was obtained. Times from the onset and peak of acute chest pain were recorded on a dedicated form for all patients. Patients with terminal kidney failure requiring regular dialysis were excluded. The study was performed according to the principles of the Declaration of Helsinki and was approved by the local ethics committees.
All patients underwent a clinical assessment that included medical history, physical examination, 12-lead ECG, continuous ECG monitoring, pulse oximetry, standard blood tests, and chest radiography. Levels of cTn were measured at presentation and serially thereafter as long as clinically indicated. Timing of the assessments and treatment of patients were left to the discretion of the attending physician.
Based on findings from the published literature and the collective clinical experiences of the investigators, we predefined 34 CPCs, including chest pain location (midchest, left side of the chest, right side of the chest, inframammillary location, and supramammillary location), size of the pain area (≥3 or <3 cm in diameter), pain quality (pressure, stabbing, burning, or aching or attendant dyspnea), radiation (to the throat, the left or the right arm/shoulder, both shoulders, back, or abdominal region or none), onset (sudden, first episode, episodes in the last 30 days, episodes >30 days ago), duration (<2, 2-30, or >30 minutes), dynamics of the pain (increasing, stable, or decreasing), aggravating or inducing factors (exertional pain aggravated by exercise, positional pain aggravated by movements, pleuritic pain aggravated by breathing and coughing, chest wall tenderness aggravated by palpation, and chest pain aggravated by emotional stress), and relieving factors, such as response to nitrates. Response to nitrates was defined as a reduction in the visual analog scale (VAS) of at least 2 cm in patients’ self-reported pain within 5 minutes of the initial dose of sublingual or spray nitroglycerin.32 Pain severity was quantified using a VAS, ranging from 0 cm for no pain to 10 cm for worst pain.33 To reflect the patient’s symptoms precisely, any combination of the predefined CPCs was possible. These characteristics were recorded in the ED during a patient interview by trained physicians and nurses on a standardized case report form with figures specifically designed to allow uniform and simple assessment of chest pain. Chest pain characteristics were uniformly recorded by physicians blinded to the 12-lead ECG and cTn levels.
Adjudication of final diagnoses was performed centrally in the core laboratory (University Hospital Basel) for all patients twice. The first method used conventional cTn levels measured on site (this method was used in the initial analyses to examine the performance of hs-cTn assays23,34,35). The second method included levels of hs-cTnT (Roche) to also take advantage of the higher sensitivity and higher overall diagnostic accuracy offered by hs-cTn assays.36,37 This process allows the additional detection of small AMIs that were missed by the adjudication based on findings of the conventional cTn assays. Two independent cardiologists (T.R. and P.H.) reviewed all available medical records pertaining to the patient from the time of ED presentation through the 90-day follow-up, including patient history, ECG, echocardiography, lesion severity, and morphological features in coronary angiography, and results of the physical examination, laboratory testing (including hs-cTnT levels), radiologic testing, and cardiac exercise testing. In situations of disagreement about the diagnosis, the cases were reviewed and adjudicated in conjunction with a third cardiologist (C.M.).
Acute myocardial infarction was defined and cTn levels were interpreted as recommended in current guidelines.38- 40 In brief, AMI was diagnosed in the presence of evidence of myocardial necrosis in association with a clinical setting consistent with myocardial ischemia. Myocardial necrosis was diagnosed by at least 1 cTn value above the 99th percentile (or for the conventional cTn assays, >10% imprecision value if not fulfilled at the 99th percentile) with a significant increase and/or decrease.39,41,42
All other patients were classified as having no AMI for this analysis. The no-AMI group included patients with unstable angina, cardiac but noncoronary disease (eg, tachyarrhythmia or perimyocarditis), noncardiac chest pain, and symptoms of unknown origin with physiological levels of hs-cTnT.
Our primary objective was to detect sex-specific CPCs that would allow more accurate differentiation of women with AMI from women with other causes of acute chest pain.18- 20 Diagnostic uncertainty was quantified using 2 different methods. First, we compared the likelihood of an acute coronary syndrome quantified by VAS by the treating physician about 90 minutes after the patient’s presentation to the ED with the adjudicated final diagnosis using the area under the receiver operating characteristics curve. Second, we quantified the incidence of disagreement between the 2 adjudicating cardiologists (requiring involvement of a third cardiologist). Continuous variables are presented as medians with interquartile ranges; categorical variables, as numbers and percentages. Continuous variables were compared with the Mann-Whitney test. Likelihood ratios (LRs) with 95% CIs were calculated to assess the specific value of each binary CPC to diagnose AMI. To assess differences in the diagnostic performance of CPCs between women and men, the interaction P value was calculated using logistic regression analysis, with sex, the specific CPC, and their interaction (ie, a combined variable representing the product of sex and CPC) entered into the model. Pain severity quantified using a VAS was expressed as a continuous variable (median [interquartile range]). Box plots for men and women with and without AMI were used to depict the sex differences. Cox proportional hazards regression analysis was used to assess the association of CPCs with mortality during long-term follow-up.
All hypothesis testing was 2-tailed, and P < .05 was considered statistically significant. All statistical analyses were performed using commercially available software (SPSS for Windows, version 21.0 [SPSS, Inc] and MedCalc, version 220.127.116.11 [MedCalc software]).
Among 2475 consecutive patients included in the study, 796 (32.2%) were women and 1679 (67.8%) were men. The adjudicated final diagnosis was AMI in 143 women (18.0%) and 369 men (22.0%) (P = .02). Among patients with AMI, 31 women and 55 men had an ST-elevation AMI (P = .43); 112 women and 314 men had a non–ST-elevation AMI (P = .004).
Table 1 provides baseline characteristics of the female and male groups and highlights statistically significant differences between women and men. These differences include being older, having a lower prevalence of some cardiovascular risk factors, and having a lower prevalence of previous coronary artery disease in women.
The prevalence of the predefined CPCs in women and men is given in Table 2. Although most CPCs were reported with a similar frequency in women and men, 11 of the predefined 34 CPCs differed significantly between women and men (P < .05). Women more often reported pressurelike pain, attendant dyspnea, pain aggravated by palpation, pain radiating to the throat or the back, sudden onset of pain, or pain duration longer than 30 minutes. In addition, women reported slightly higher pain severity. Women less often reported no radiation, radiation to the right side, and pain duration of 2 to 30 minutes.
The prevalence of the predefined CPCs in women and men with an adjudicated diagnosis of AMI is given in the Supplement (eTable 1). In the AMI group, most CPCs were reported with similar frequency in women and men, whereas 5 of the predefined 34 CPCs differed significantly between women and men. These CPCs were related to pain radiation and pain duration/dynamics. Women with AMI reported less frequent radiation to the right arm/shoulder, duration of pain from 2 to 30 minutes, and decreasing pain; they reported more frequent radiation to the back and duration of pain longer than 30 minutes. We tabulated these data separately for ST- and non–ST-elevation AMI (Supplement [eTables 2 and 3]).
Thirty-one of 34 CPCs (91.2%) showed similar LRs for the diagnosis of AMI in women and men. Most CPCs did not discriminate AMI from other causes of acute chest pain in women or men; they had LRs close to 1 and rather large 95% CIs (Figure 1). However, the following CPCs significantly decreased the likelihood of the diagnosis of AMI, with similar LRs in women and men: stabbing pain; aggravation of the pain by breathing, movement, or palpation; pain located in the left side of the chest and inframammillary pain; pain without radiation; and pain duration of less than 2 minutes.
The following CPCs significantly increased the likelihood of the diagnosis of AMI, with similar LRs in women and men: pain aggravated by exertion or relieved by nitrates, pain location on the midchest and/or in the right side of the chest, pain area larger than 3 cm, radiation to the left and/or right shoulder/arm, and radiation to both shoulders (Table 3). More severe pain (quantified using a VAS) increased the likelihood of the diagnosis of AMI in both sexes (Figure 2).
Only 3 CPCs (8.8%) related to pain duration and pain dynamics showed a sex-specific diagnostic performance with statistically significantly different LRs between women and men. Pain duration of 2 to 30 minutes decreased the likelihood of an AMI diagnosis in women but increased the likelihood in men (P = .01 for the difference between women and men). Pain duration exceeding 30 minutes increased the likelihood of an AMI diagnosis in women but was neutral in men (P = .04 for the difference between women and men). Decreasing pain intensity tended to decrease the likelihood of an AMI diagnosis in women but tended to increase the likelihood in men (P = .04 for the difference between women and men) (Table 3 and Figure 2).
Likelihood ratios of all possible combinations of CPCs in women and men that were significantly associated with AMI in univariate analysis are tabulated in the Supplement (eTables 4 and 5). We found no statistically significant interaction for any of the combinations of CPCs with sex.
The combination of a patient’s history, including a detailed assessment of CPCs, the 12-lead ECG findings, and results of cTn testing, form the diagnostic cornerstones for the early diagnosis of AMI.5,6 In this prospective multicenter study, we assessed 34 predefined and uniformly recorded CPCs and aimed to contribute to an improvement in the management of suspected AMI in women by exploring the diagnostic value of sex-specific CPCs. Sex-specific CPCs would allow physicians in the ED to distinguish women with AMI from women with other causes of acute chest pain with greater accuracy. We report 3 major findings. First, although most of the CPCs were reported with similar frequency in women and men, 11 of the predefined 34 CPCs were reported to be significantly different in women and men. This observation from our large cohort with acute chest pain extends and corroborates findings from previous AMI cohorts.20- 22 These previous studies were unable to address the key question whether attention to any of these CPCs would improve the early diagnosis of AMI in women. Second, most of the assessed CPCs did not differentiate AMI from other causes of acute chest pain, with LRs close to 1 and rather large 95% CIs. By validating the findings from smaller pilot studies8,43- 45 and the guidelines recommendations,4,6 our data confirm that CPCs are not powerful enough to be used as a single tool in the diagnosis of AMI and need to be used always in conjunction with the ECG and cTn test results in the diagnosis of AMI. Third, 31 of 34 CPCs (91.2%) showed similar LRs for the diagnosis of AMI in women and men, and only 3 CPCs (8.8%) showed a sex-specific diagnostic performance. In further support of the concept of sex differences in the perception of time,33 these CPCs were related to pain duration and dynamics. However, because their LRs were close to 1 with rather wide 95% CIs, these CPCs did not seem clinically helpful. In addition, we must acknowledge that 3 of 34 CPCs showing a statistically significant interaction with sex may also have been the result of chance.
The findings of this study regarding sex-specific CPCs extend and corroborate current efforts to consider sex-specific criteria of the 3 diagnostic pillars in the early diagnosis of AMI.4,32 First, the criteria regarding the extent of ST-segment elevation required in the 12-lead ECG in V2 and V3 to qualify for an ST-segment elevation AMI in current guidelines are slightly different in women and men.4 These sex-specific ECG criteria still await prospective validation in consecutive patients with acute chest pain. Second, recent studies using hs-cTn assays indicate that women have significantly lower 99th percentile values of a healthy reference population compared with men.34 This observation has initiated a discussion about the advantages and disadvantages of using a sex-specific 99th percentile as the clinical cutoff for cTn values.32 Data from large multicenter studies will need to define the clinical merit of such a novel approach.
Some limitations of our study merit consideration. First, our findings are based on 34 predefined CPCs and attendant dyspnea. We did not uniformly assess all possible accompanying symptoms, such as nausea, vomiting, and sweating, and did not enroll patients without chest pain. Because about 8% of patients with AMI seem to present without chest pain,1 this small but important subgroup requires future study. In theory, symptoms other than chest pain may provide sex-specific diagnostic value. We consider this possibility rather unlikely because the LR of attendant dyspnea for AMI was similar in women and men. Second, the patients undergoing evaluation in our study presented with acute chest pain to the ED. We do not know whether our findings can be extrapolated to patients presenting to primary care settings. Third, we cannot generalize these findings to patients with terminal kidney failure requiring dialysis because they were excluded from our study. Fourth, the adjudication of the final diagnosis was based on all available information, including CPCs. The study methods might have introduced a small bias in favor of typical CPCs. However, given the overall dismal performance of CPCs in this study, we consider any possibly remaining bias to be minimal.
In conclusion, differences in the sex-specific diagnostic performance of CPCs overall are small. Our findings do not seem to support the use of CPCs specific to women in the early diagnosis of AMI in women.
Accepted for Publication: September 16, 2013.
Corresponding Author: Christian Mueller, MD, Department of Cardiology, University Hospital Basel, Petersgraben 4, CH-4031 Basel, Switzerland (firstname.lastname@example.org).
Published Online: November 25, 2013. doi:10.1001/jamainternmed.2013.12199.
Author Contributions: Drs Rubini Gimenez and C. Mueller had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Rubini Gimenez, Reiter, Twerenbold, Reichlin, C. Mueller.
Acquisition of data: Rubini Gimenez, Reiter, Twerenbold, Reichlin, Wildi, Haaf, Zellweger, Hoeller, Moehring, Sou, M. Mueller, Meller, Stallone, Henseler, Bassetti, C. Mueller.
Analysis and interpretation of data: Rubini Gimenez, Reiter, Reichlin, Wildi, Zellweger, Denhaerynck, Meller, Stallone, Geigy, Osswald, C. Mueller.
Drafting of the manuscript: Rubini Gimenez, Reiter, Twerenbold, Reichlin, C. Mueller.
Critical revision of the manuscript for important intellectual content: Rubini Gimenez, Reiter, Twerenbold, Reichlin, Wildi, Haaf, Zellweger, Hoeller, Moehring, Sou, M. Mueller, Denhaerynck, Meller, Stallone, Bassetti, Geigy, Osswald, C. Mueller.
Statistical analysis: Rubini Gimenez, Reiter, Wildi, Zellweger, Denhaerynck, Meller, Stallone, C. Mueller.
Obtained funding: Osswald, C. Mueller.
Administrative, technical, and material support: Rubini Gimenez, Twerenbold, Wildi, Hoeller, Sou, Stallone, Henseler, Osswald, C. Mueller.
Study supervision: Rubini Gimenez, Reichlin, Haaf, Denhaerynck, Osswald, C. Mueller.
Conflict of Interest Disclosures: Dr Reichlin has received research grants from the Swiss National Science Foundation (PASMP3-136995), the Swiss Heart Foundation, the University of Basel, the Professor Max Cloetta Foundation, and the Department of Internal Medicine, University Hospital Basel; he has received speakers’ honoraria from Brahms and Roche. Dr C. Mueller has received research grants from the Swiss National Science Foundation, the Swiss Heart Foundation, the Cardiovascular Research Foundation Basel, 8sense, Abbott, ALERE, Brahms, Critical Diagnostics, Nanosphere, Roche, Siemens, and the University Hospital Basel; he has received speakers’ honoraria from Abbott, ALERE, Brahms, Novartis, Roche, and Siemens. No other disclosures were reported.
Funding/Support: This study was supported by research grants from the Swiss National Science Foundation, the Swiss Heart Foundation, the Cardiovascular Research Foundation Basel, the University of Basel, and the University Hospital Basel.
Role of the Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank the patients who participated in the study, the staff of the participating EDs, and the research coordinators. Michael Freese, RN, Claudia Stelzig, MS, Esther Garrido, RN, Irina Klimmeck, RN, Melanie Wieland, RN, Janine Voegele, RN, and Beate Hartmann, PhD, provided valuable efforts. Fausta Chiaverio assisted as laboratory technician. None of these individuals received financial compensation.
Correction: This article was corrected on December 4, 2013, to reformat a reference.