Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Getting better - myocardial infarction

Study
Results
Comment

People often ask the question about whether evidence-based medicine “works”. This is a difficult one, for several reasons. The question as it stands is indistinct, because we all use evidence of a sort. Perhaps it should be whether using good evidence is better than using bad evidence, or no evidence. Philosophers, professional and amateur, could spend ages over the question before even attempting an answer.

Again, we tend to think of EBM as a result about a particular intervention, achieved usually through a systematic review (and probably some form of meta-analysis), or with results from a solid randomised trial. Yet healthcare is multidimensional, and often involves complex packages of care, of which a single intervention may play only a small part.

Bandolier 100 highlighted a study from South Derbyshire showing that over five years from 1995 to 1999 mortality over 30 days and one year after a heart attack showed consistent year-on-year reductions, alongside improvements in the use of treatments for which there was a strong evidence base. Bandolier had overlooked a US report showing the same thing, but over a longer period [1].

Study

The review used data from a variety of sources, including population-based studies reporting at least 10 years of data to determine changes in intervention rates for different therapies, meta-analyses of randomised trials to estimate benefit, and incidences of myocardial infarction in the US from a national hospital discharge survey. From these, a 30-day mortality was calculated, and the contribution of various treatment changes calculated.

Results

The main results are shown in Table 1. The age and sex adjusted incidence of myocardial infarction fell between 1975 and 1995 by 29%, with most of the reduction in the early 1990s. More people had hypertension. Over the period the average age of patients with myocardial infarction increased by five years, with 7% more women diagnosed.


Table 1: Main changes in myocardial infarction incidence and treatment in the USA


Variable
1975
1980
1985
1990
1995
New MI (age/sex adjusted to 1996, thousands)
613
527
591
481
437
Patient characteristics and infarct severity
Age (years)
64
66
68
68
69
Hypertension (%)
40
48
48
51
55
Anterior or lateral location (%)
49
49
48
41
35
Q-wave infarction (%)
73
67
61
55
48
Use of therapies with clear mortality benefit
Aspirin (%)
5
6
30
62
75
Beta-blockers (%)
21
42
48
47
50
Thrombolytics (%)
0
0
9
25
31
ACE inhibitors (%)
0
0
0
13
21
Primary angioplasty (%)
0
0
0
NA
9
30-day fatality rate (age/sex adjusted to 1995, %)
27
28
24
20
17


Severity of acute myocardial infarction seems to have declined. Though this may be because of better diagnostic sensitivity, it also indicates that significant population changes occurred over the period, probably relating to adherence to healthier lifestyles.

Using therapies with clear reduction of mortality after a heart attack increased over the period (Table 1). Aspirin use increased from 5% in 1975 to 75% in 1995, for instance. The result was a consistent fall in 30-day mortality from just under 30% in 1975 and 1980, to 17% in 1995. (Table 1, Figure 1). The largest benefits were calculated to come from use of aspirin (30% contribution) and thrombolysis (15%).


Figure 1: Age and sex adjusted 30-day mortality after myocardial infarction in the USA
















Comment

This report ties in well with that from South Derbyshire, which reported a 13% 30-day mortality by 1999. Both reports show increased use of interventions with a good evidence base in a complex situation combining to deliver large reductions in mortality after a heart attack.

The US study also indicates that improved diagnosis, contributing about 18% to the reduced mortality, mainly through increased use of cardiac enzyme tests. Even better tests now coming into use (at least in the US) could contribute further by diagnosing smaller infarcts, so that treatment can be appropriate.

As well, there is the contribution of healthier lifestyles. Fewer heart attacks, and less severe heart attacks, both contribute to reduced mortality. Again, here is an area with a good evidence base (just look at the Bandolier Internet section on healthy living for an overview). Part of the US population has clearly got the message.

So it may be complex, and getting it all into practice may be spotty, but over all, in behaviour, diagnosis, and treatment, use of good evidence seems to be producing the expected benefits.

Reference:

  1. PA Heidenreich, M McLellan. Trends in treatment and outcome for acute myocardial infarction: 1975-1995. American Journal of Medicine 2001 110: 165-174.

previous or next story