Options recommend for Ann 

The percentages of respondents who indicated that they would recommend the options listed in the case study were, in order of popularity:

There were significant correlations between believing that HRT was beneficial and recommending HRT but not recommending physical activity, a Mediterranean diet, a beta-blocker or a “statin”.

A standard statistical equation using respondents’ beliefs about the impact of HRT on fractures, cardiovascular risk and breast cancer correctly predicted only 70% of respondents’ decisions to recommend HRT or not.[1]  This suggests that despite the fact that those 3 issues were the most important for Ann, 30% of respondents recommendations were swayed by other factors unknown to us.  Respondents’ beliefs about the impact of HRT on thromboembolism and gallbladder disease did not add significantly to the prediction. 

Many respondents wanted more information before making recommendations.  (NPS Case Studies are currently limited to one page to make them easier for busy GPs to complete.)  Many commented that their recommends for “statins”, ACE inhibitors and HRT would depend on Ann’s LDL cholesterol, blood pressure and bone density test results respectively.  Respondents also wanted to check on smoking, alcohol, caffeine, calcium, diabetes, screening mammograms and cervical cytology. 

Some respondents commented that they would inform Ann about the “pros and cons” of HRT to enable her to make her own decision.  (However it is clear that different doctors would tell Ann quite different things about the pros and cons of HRT.)

Our comment: 

We would recommend a Mediterranean diet, physical activity, aspirin, and a beta-blocker.  If Ann’s total cholesterol level was >5 we would also recommend a “statin”. 

Explanation:

The challenge is to get the right balance of RCT evidence, observational evidence, clinical experience and understanding of the unique individual patient so as to provide the best advice.  Ideally we would like to see at least 2 RCTs involving large numbers of women like Ann who have had a myocardial infarct.  In the case of diet and physical activity alone few trials have been done because funding such trials would not be profitable for drug companies.

One RCT, the Lyon heart diet study, has shown large benefit from a “Mediterranean” diet (more bread, more vegetables, more fruit, more fish, and less meat) although the report did not mention how many women participated.[2]  Observational evidence from the Nurses Health Study supports the belief that good diets are very beneficial for women.[3]  It can be difficult to improve diet and physical activity levels if patients are not motivated.  However the results of the Lyon heart diet study suggest that if patients are motivated (eg by personal experience of a myocardial infarct) then greater benefits can be achieved with lifestyle changes than could be expected with medication. 

We could not find any trials of exercise alone post myocardial infarction but 4 RCTs have found that cardiac rehabilitation that includes physical activity is beneficial.[4]^,[5],[6],[7]  The Nurses Health Study also supports the belief that physical activity is very beneficial for women.[8]

Meta-analyses of many trials involving men and women have found that aspirin[9][10] and beta-blockers[11] are beneficial post myocardial infarction. 

RCTs have shown benefit from “statins”[12] for people who have had a myocardial infarct even if their lipid levels are normal although the benefit may be less if the total cholesterol level is less than 5 mmol/L.  Patients with existing coronary heart disease are eligible for Pharmaceutical Benefit Scheme subsidised lipid lowering drugs if their total cholesterol is > 4 mmol/L. 

RCTs have shown benefit from ACE inhibitors but the magnitude of benefit is small so we have given it lower priority.[13]  If Ann had high blood pressure not controlled by a beta-blocker then an ACE Inhibitor would be a good choice. 

The hypothesis that traditional Japanese diets are superior to other diets and the hypothesis that isoflavins are better than placebo for the outcomes of interest here have not yet been properly tested.

HRT would reduce Ann’s hot flushes but she has not requested treatment for them.  For women like Ann who have had a myocardial infarct, HRT has been shown to be potentially harmful in RCTs of up to 4.1 years duration.  The possibility that longer-term use may do more good than harm has not been excluded but any late benefit would have to be very large to make up for the early harm.  The best evidence available to date indicates increased thromboembolism and maybe gallbladder disease, possible early cardiovascular harm with no medium term cardiovascular benefit, and uncertainty with regards to impact on breast cancer and fracture.  On balance, the risks of harms outweigh the uncertain benefits.  We are also concerned that use of HRT may also be harmful by diverting attention from more beneficial interventions.