An Assessment of the Cardiovascular Effects
of Hormone Replacement Therapy
on Post-menopausal Hyperlipidemic Cardiac Patients
Treated with HMG-CoA Reductace Inhibitors.
Ruthmarie Shea Atallah R.N.
Pierre Charbel Atallah
Rochester Medical Center, P.C.
HMG-CoA Reductace inhibitors (statins) have been shown to be effective in reducing the frequency of coronary incidents, including cardiac death, in hyperlipidemic patients. It is currently thought that hormone replacement therapy (HRT) in post-menopausal women can reduce coronary incidents. This study seeks to determine whether there is a difference in cardiovascular outcomes, including fatal myocardial infarctions, non-fatal myocardial infarctions, cerebral vascular accidents, transient ischemic attacks, and necessity for surgical interventions limited to: Coronary Artery Bypass Graphs, Percutaneous Transluminal Coronary Angiography and Stent implantation, between post-menopausal, hyperlipidemic, cardiac patients who are treated by statins alone, and those treated with statins in combination with HRT.
A retrospective analysis was conducted to determine whether statistically significant differences in outcomes could be identified in the group treated with HRT. The P value of 0.1676 fails to give evidence against the null hypothesis. Post-menopausal, hyperlipidemic, women with cardiac pathophysiology in this study did not exhibit statistically significant differences in morbidity or mortality when treated with HRT.
Sample Design-Randomization: Because of the importance of sampling groups within a large population separately, a stratified random sample was used. First, a large population (n=488) of females who fit the general criteria of being treated with statins for hypercholesterolemia was selected. Next, those patients under the age of 60 were eliminated. This was an intentional part of the design, because there is general acceptance within the medical community that, perhaps as a result of endogenous estrogens, women do not exhibit adverse cardiac events until their sixties. To eliminate this confounding factor, those under 60 years of age were excluded from the study.
The remaining population was then divided into two groups of similarly situated females. These strata were: 1) post-menopausal hyperlipidemic female cardiac patients who were being treated with statins alone, and not receiving HRT, and 2) post-menopausal hyperlipidemic female cardiac patients who were being treated with statins and also with HRT. Separate Simple Random Samples (SRS) were obtained for each stratum.
The SRS were obtained by labeling all individuals in the population by coded eight-digit number. A random number generator of the TI-83 graphing calculator was used to obtain the necessary sequences of random numbers between 0 and 9. To further randomize, the digits from the end of the coded number were then matched.
Forty study participants were selected. Twenty were assigned to each of the two populations. Group I ( no HRT) had a mean of 70.1 years of age with a median age of 70. The treatment and response of this group was followed for a mean period of 59.25 months. Group II (HRT) had a mean age of 68.35 years, with a median age of 76.
The thinking was that if treatment with statins is beneficial in avoiding adverse cardiac outcomes, and HRT therapy is also beneficial in avoiding adverse cardiac outcomes, then perhaps treating with both pharmacological agents, as combination therapy, would yield even better cardiac outcomes. Data analysis revealed, however, that combination therapy did not improve outcomes in this group. It appeared that those who had dual therapy, statins and hormone replacement therapy combined, did worse that those treated with statins alone. The initial finding was that those who were on both therapies (Group II) underwent cardiac surgery at twice the rate of those treated with statins alone.
The two-sided hypothesis is as follows:
Ho: p1 = p2
Ho: p 1 ¹ p 2
In this study p1 represents the proportionate incidence of adverse coronary events in Group I, p2 represents the proportionate incidence of adverse coronary events in Group II. The null hypothesis states that there is no difference in adverse coronary outcomes between Group I and Group ll. The alternative hypothesis is that there is a difference in outcome between Group I and Group II.
Adverse coronary events were defined as morbidity and mortality. For purposes of this study morbidity was defined as an experience of Cerebral Vascular Accident/ Stroke (CVA), Transient Ischemic Attack (TIA) Myocardial Infarction (MI) or surgical intervention limited to; Coronary Artery Bypass Graphs (CABG), Percutaneous Transluminal Coronary Angiography, (PTCA) or Stent Implantation. Mortality was defined as death.
Twenty percent of Group I required surgery to the coronary arteries. There were 0 incidents of cardiac death, 0 incidents of non-fatal MI, 0 incidents of CVA/TIA, and 4 incidents of surgical intervention defined as CABG, PTCA or Stent implantation.
Forty percent of Group ll required surgery to the coronary arteries. There were 0 incidents of cardiac death, 0 incidents of non-fatal MI, 0 incidents of CVA/TIA, and 8 incidences of surgical intervention, defined as CABG, PTCA or Stent implantation.
When success is defined as the absence of surgery, Group I experienced an 80% success rate, while Group II experienced only a 60% success rate. The difference between the mean of the two groups was 0.20. The variance of the difference was 0.02. The standard deviation of the difference was 0.1414. The standard error of the difference was 0.1414. The 95% Confidence Interval is (-) 0.0771, 04771. As can be seen, zero is within this interval. It can be concluded, therefore, that there is no significant difference between the two groups. While Group I is 20% more likely to avoid surgery, this success is without statistical significance. This assertion is made with 95% Confidence and a margin of error of 14.14%.
Significance testing was conducted to test the null hypothesis that p1=p2.
A Z –test was performed in which p1 was subtracted from p2
and divided by the standard error of the pooled difference. The Z test statistic
was 1.38 which correlates with a Table A entry of 0.9162. After calculating for
a two-sided hypothesis, P is 0.1676. High P values do not provide evidence
against the null hypothesis. This indicates that even if H:o is true, a value as
extreme a Z will be observed approximately 17% of the time. Using a fixed alpha
test at level alpha 0.05 does not provide evidence against the null hypothesis.
A fixed alpha test would not reject Ho until 0.20
The subjects in this study did not exhibit statistically significant differences in morbidly or mortality when treated with Hormone Replacement Therapy. Further, the fixed alpha test did not reject that null hypothesis at 0.05, and would not reject Ho until 0.20. The P value of 0.1676 fails to give evidence against the null hypothesis. Therefore, the study suggests that the administration of Hormone Replacement Therapy (HRT) to post-menopausal women with hyperlipidemia and concomitant cardiac disease is of no statistical benefit in reducing morbidity or mortality from adverse cardiovascular outcomes.