Gender and Chest Pain Evaluation, Silbergleit, McNamara 115

I PRELIMINARY REPORTS

Effect of Gender on the Emergency Department Evaluation of Patients with Chest Pain

Robert Silbergleit, M D , Robert M . McNamara, M D

I ABSTRACT

Objective: To assess chest pain evaluation as reflected in the docu- mentation of the evaluation process for women vs men in one emer- gency department (ED).

Methods: In this retrospective case series, patient charts were re- viewed for documentation in accordance with a clinical policy for chest pain evaluation. Patient demographics and the frequencies of inclusion of the following items were determined: five descriptors of chest pain, associated symptoms, risk factors for coronary artery dis- ease, receipt of physical examination, and receipt of ECG.

Results: Over a three-month period, 132 men and 150 women were evaluated for chest pain and entered in the study. There was no significant difference in age between the men and the women overall, but in the subgroup of patients who were admitted to the hospital, the women were significantly older than the men by an average of five years (p = 0.04). Fifty-five percent of all the patients were admitted to the hospital. The men were admitted to the hospital significantly more often than were the women (p = 0.01), with a relative risk of admission for women vs men of 0.76 (95% CI = 0.62- 0.94). There was no significant difference between the men and the women for any of the process of evaluation items in the overall group or in the hospital-admission and release-home subgroups.

Conclusion: The authors’ findings do not support the existence of a gender difference in ED chest pain evaluations, as reflected by doc- umentation of the evaluation process. However, men were more likely to be admitted to the hospital for evaluation of coronary artery disease than were women.

Acad. Emerg. Med. 1995; 2:115-119.

~

116 ACADEMIC EMERGENCY MEDICINE FEB 1995 VOL 2/NO 2

1 Coronary artery disease is the single largest killer of both men and women in the United States. Overall, women receive one third of the 4.5 million new diag- noses of coronary artery disease each year; however, the incidence and prevalence of coronary artery disease among postmenopausal women are nearly the same as those among similarly aged men.’ Unfortunately, the misconception that heart disease is predominantly a dis- ease of men is widespread among both the public and medical professionals. Multicenter studies such as those by the Survival and Ventricular Enlargement Group have shown that women with identified coronary artery disease are treated less aggressively than are men, de- spite higher degrees of cardiac di~abili ty.~.~ Gender bias also has occurred in coronary artery disease research. Many of the most important and influential studies of diagnosis, prevention, and intervention of heart disease have excluded women.4 The NIH recently announced the Women’s Health Initiative to correct research bias in cardiovascular disease and other areas.

Gender bias in the evaluation and diagnosis of cor- onary artery disease has not been well studied, but some physicians have asserted in the lay press that women presenting to emergency departments (EDs) with chest pain are not taken seriously regarding their potential for myocardial infarction (MI).s There has been no study specifically addressing the evaluation received by women presenting to the ED with chest pain.

This study addresses whether men and women com- plaining of chest pain are equally assessed by well-trained emergency physicians (EPs), as reflected by documen- tation of the evaluation process in the ED. The Amer- ican College of Emergency Physicians (ACEP) clinical policy guidelines for patients presenting with chest pain6 were used as uniform reference criteria to compare the processes of care for men and women.

I METHODS

Study Design We conducted a retrospective analysis of ED record

documentation for men and women who had presented with chest pain. Documentation of explicit criteria (listed below) was chosen as our measure of patient evaluation, because documentation of these items suggests an ad- equate process of care as denoted by the development of an adequate information base with which to make decisions.’ This assessment method is commonly used in the field of quality

Population Base and Subjects Study subjects presented to the ED of the Miseri-

cordia Division of Mercy Catholic Medical Center, an inner-city community teaching hospital, over a three-

month period (January 1 to March 31, 1990). The an- nual number of patient visits to this E D is approximately 28,000. Inclusion criteria included either a chief com- plaint of chest pain and ED release with any diagnosis or hospital admission with a cardiac diagnosis to an ECG-monitored bed. Exclusion criteria included either hospital admission with a noncardiac diagnosis, age less than 30 years, or a history of recent traumatic injury to the chest.

The treating physicians were all residents and faculty of an emergency medicine (EM) residency program. The residents were usually in the second or third year of training. Thirteen of the 16 EM faculty providing coverage during the study period were residency-trained and diplomates of the American Board of Emergency Medicine. Of the remaining three faculty, two were board-certified in EM and at least one other specialty, and one was board-certified in internal medicine alone.

Measurements Documentation of items on a clinical policy guide-

line for treatment of adult patients presenting with a chief complaint of chest pain and without a history of trauma6 was determined through retrospective ED chart review. Data collected were patient-identifying infor- mation, age, gender, and the frequency of physician documentation of items from the guidelines: five de- scriptors of chest pain (onset, location, duration, char- acter, and radiation); associated symptoms (shortness of breath, diaphoresis, nausea, and vomiting); risk fac- tors for coronary artery disease (previous cardiac dis- ease, hypertension, diabetes, family history,, smoking history, and elevated level of cholesterol), receipt of physical examinations, including a cardiovascular and chest examination, and receipt of ECG. This retro- spective review was not contemplated during the study period; therefore, all the physicians who performed the patient evaluations were blinded to the study. Data collection was carried out by the investigators using a standardized data form.

Data Analysis Data were compiled, and the continuous variables

of age and documented number of descriptors of chest pain were compared by two-sided t-test for all patients and for the admitted and released patient subgroups. Categorical variables, including hospital admission, documented associated symptoms, risk factors, fully documented physical examination, and ECG order, were compared by chi-square analysis. Relative risk of ad- mission for women vs men was determined without ad- justment for actual workup findings (e.g., ECG ab- normalities, character of pain) and a 95% CI was calculated.

Gender and Chest Pain Evaluation, Silbergleir, McNamara 117

I TABLE I Documentation for All Patients, by Gender ~ -

Men Women ( n = 132) (11 = 150) p-va 1 ue

Age-mean & SD

Number of chest pain descriptors-mean t SD

Number having associated symptoms

Number having coronary artery disease risk factors

Number receiving cardiovascular and chest examinations

Number receiving ECGs

Number admitted to the hospital

-

54 t 16yr

3.1 t 1.6

106 (XI%,)

88 (67%)

122 (93%)

I16 (89%)

83 (63%)

55 5 16yr 0.70

3.2 k 1.4 0.52

124 (83%) 0.70

114 (76%) 0.10

141 (94%) 0.77

134 (89%) 0.X3

72 (48%) 0.01

Our study was conducted with a confidence of 95% (5% a-error), and it had a power of 90% (10% p-error) to detect less than a 10% difference from 100% doc- umentation of the items suggested by the clinical policy guidelines6 or to detect an 18% difference, from an 80% complete documentation baseline, between the gender groups for the categorical measures.

I RESULTS

The data for all the patients (Table 1) and for the ad- mitted (Table 2) and released (Table 3) subgroups are shown. Two hundred eighty-two patients, 132 men and

.....................................................................................

150 women, were entered in the study. The mean age was 54 years ( 2 16 SD). There was no significant dif- ference in age between the men and the women overall, but in the subgroup of patients who were admitted to the hospital, the mean age of the women was signifi- cantly higher than that of the men, by five years (p = 0.04). There was no significant difference in the ages of the released men and women. Fifty-five percent of all the patients were admitted to the hospital. The men were admitted significantly more often than were the women (p = 0.01). with a relative risk of admission for women vs men of 0.76 (95% CI = 0.62-0.94).

There was no significant difference between the men

I TABLE 2 Documentation for Patients Admitted to the Hospital, by Gender ..............................................................................................................................................................................................................................................

Men Women ( n = 83) (n = 72) p-value

Age-mean t SD 59 2 15 yr 64 2 13yr 0.04

Number of chest pain descriptors-mean +. SD 3.1 -r- 1.5 3.2 f 1.3 0.86

Number having associated symptoms 65 (79%) 55 (76%) 0.67

Number having coronary artery disease risk factors 53 (65%) 48 (67%) 0.79

Number receiving cardiovascular and chest examinations 76 (93%) 65 (90%) 0.69

Number receiving ECGs 80 (98%) 71 (99%) 0.64

I TABLE 3 Documentation for Patients Released from the Emergency Department, by Gender

Men (I1 = 49)

Women (I7 = 78) pvulue

Age-mean ? SD

Number of chest pain descriptors-mean k SD

Number having associated symptoms

Number having coronary artery disease risk factors

Number receiving cardiovascular and chest examinations

Number receiving ECGs

45 t 13 yr 46 k 14 yr 0.64

3.1 5 1.6 3.3 t- 1.4 0.53

41 (84%) 69 (88%) 0.44

35 (71%) 66 (85%) 0.07

46 (94%) 76 (97%) 0.32

35 (73%) 63 (XI%,) 0.33

118 ACADEMIC EMERGENCY MEDICINE FEB 1995 VOL 2/NO 2

and the women for any of the process of evaluation items in the overall group or in either subgroup. As identified in Tables 1 through 3, similar numbers of descriptors of chest pain, associated symptoms, and risk factors were recorded for the men and women patients. Cardiovascular and chest examinations were usually re- corded for patients of both genders. Nearly all the ad- mitted patients had documented ECGs, as did 81% of the women and 73% of the men released from the ED (p = 0.33).

I DISCUSSION ...............................................................................................................

These data show that the evaluation documented for patients presenting with nontraumatic chest pain in our ED is not influenced by the gender of the patient. Be- cause we sought to assess whether women with chest pain were not taken seriously as cardiac patients, we measured the process of care received in the E D rather than outcome measures. Process of care assessment is a common quality assurance tool and reflects the eval- uation a patient receives.

The men in our study were admitted to the hospital significantly more often than were the women. On the surface this implies a more conservative bias in the dis- position of men chest pain patients, but our relative risk for admission of women vs men of 0.76 did not account for the presence of positive and negative findings on clinical evaluation. Further data are required to deter- mine whether an admission bias exists. If there were a bias against admitting women in our study, the relative risk of coronary artery disease for the women vs the men in our population would have to be higher than the relative risk of hospital admission.

Unfortunately, we do not know the true relative risk of coronary artery disease for the women vs the men in our patient population. The relative risk of coronary artery disease for women vs men in the general U.S. population, however, is stated to be 0.60,’ suggesting that our admission ratio of 0.76 does not represent a bias against women. The tendency of the physicians in this study to admit older women is consistent with the results of studies reporting that women with coronary artery disease have a higher mean age than do men with the disease. 1.12 Whether this age difference for admis- sion between men and women was appropriate cannot be determined from our study.

Our results are consistent with those of previous studies that found no gender bias in the rates of un- recognized MI for which the patient was sent home from the ED,13 or in the rates of clinical diagnosis of angina preceding an index MI.’ These studies and others have attributed worsened outcomes for women with coronary artery disease not to misdiagnosis but to a combination of problems, including the more advanced age of women

with heart disease; gender bias in the underuse of an- giography, angioplasty, and bypass grafting; and rela- tively poorer results from these interventions for women compared with men.*.’

I LIMITATIONS AND FUTURE QUESTIONS ...............................................................................................................

This study is limited in scope. It studied the evaluation of women in only one ED staffed by the faculty and housestaff of an E M residency program. The results, therefore, may not reflect the treatment of women else- where or by EPs with different backgrounds. One ele- ment of the evaluation process not studied was the speed with which the evaluation was conducted. Speed of eval- uation may be critical in the subset of patients with ongoing MI, especially if thrombolytic agents are to be used. Jackson et al. described preliminary results sug- gesting that women receive thrombolytic agents in the ED approximately 25 minutes later than do men.I4

Finally, although the contents of the evaluations were not different between genders, we studied only the ac- tions of the physicians. We did not measure the preas- sessment beliefs of the physicians or the perceptions of the patients. The concern that women with chest pain are not taken seriously may reflect a differential sense of concern projected by the physicians, rather than a real difference in the extents of evaluation or the de- grees of care. These issues should be addressed in future studies.

Future studies of this subject also should involve larger populations in multiple centers and should ex- amine the effect of gender on both the evaluation pro- cess and the use and timing of ED interventions in true acute cardiac disease.

I CONCLUSION

In this study, the men and women evaluated in the E D with complaints of chest pain received the same degree of clinical evaluation, as measured by care documen- tation. We found no evidence to support the extension of the gender bias observed in the treatment of women with coronary artery disease15 to the process of initial evaluation and diagnosis in the ED.

............................................................................................

4 REFERENCES

1. Centers for Disease Control. Coronary heart disease incidence, by sex. MMWR. 1992; 41526-9.

2. Ayanian JZ, Epstein AM. Differences in the use of procedures between women and men hospitalized for coronary heart disease. N Engl J Med. 1991; 325:221-5.

3. Steingart RM, Packer M, Hamm P, et al. Sex differences in the management of coronary artery disease. N Engl J Med. 1991;

4. Healy B. The Yentl syndrome. N Engl J Med. 1991; 325:274-5. 5. Jaroff L. The biggest killer of women: heart attack. Time. 9 Nov

325:226-30.

Gender and Chest Pain Evaluation, Silbergleit, McNamara 119

6.

7.

8.

9.

10.

11.

1992; 140(19):72-3. 12. American College of Emergency Physicians. Clinical Policy for Management of Adult Patients Presenting with a Chief Complaint of Chest Pain, with No History of Trauma. Dallas, TX: ACEP, 1990. Payne BC. The medical record as a basis for assessing physician competence. Ann Intern Med. 1979; 91:623-9. Donabedian A. Criteria and standards for quality assessment and monitoring. QRB. 1986; 12:99-108. Flint LS. Hammett WH, Martens K. Quality assurance in the emergency department. Ann Emerg Med. 1985; 14:134-8. Whitcomb JE, Stueven H, Tonsfeldt D, et al. Quality assurance in the emergency department. Ann Emerg Med. 1985; 14:1199- 204. McNamara RM, Kelly JJ. Impact of an emergency medicine residency program on the quality of care in an urban community hospital emergency department. Ann Emerg Med. 1992; 21:528- 33. ....

13.

14.

15.

Cunningham MA, Lee TH, Cook EF, et al. The effect of gender on the probability of myocardial infarction among emergency department patients with acute chest pain: a report from the Multicenter Chest Pain Study Group. J Gen Intern Med. 1989;

Lee TH, Rouan GW. Weisberg MC, et al. Clinical characteristics and natural history of patients with acute myocardial infarction sent home from the emergency room. Am J Cardiol. 1987; 60:219- 24. Jackson RE, Anderson W, Peacock WF, Wilson AG. Gender bias and thrombolytic therapy [abstract]. Ann Ernerg Med. 1993; 22:942. Kuhn FE, Rackley CE. Coronary artery disease in women: risk factors, evaluation, treatment, and prevention. Arch Intern Med.

4:392-8.

1993: 153:2626-36.

Reflections .

I INVOCATION We rejoice that your embrace shepherds our deeds As we respond to those whose vital signs evoke The controlled panic of a prehospital stampede. We’re appreciative that you’ve spared the death stroke For singular patrons who’ve attempted a discharge From tri-monogrammed A’s of DK, CV, MV, or AA. All of our graduates have had their databases enlarged; The aggregate will weather the boards ultimate assay. You cajoled a minority to a preeminent scale. We’re conscious that you sustained the cures Of those who willed that common good prevail. And, gathered tonight, we beseech their effectiveness endures.

Amen

JONATHAN SINGER, MD Vice-Chair and Program Director Wright State University School of Medicine Department of Emergency Medicine Kettering, Ohio