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

Smoking with CHD

Systematic review

All of us deal with risk every day. If you drive on the roads in the UK, you accept a (roughly) 1 in 18,000 chance of dying in a road accident in one year. Given the volume of traffic on our roads, that is a risk most of us are willing to accept. But it is not an unthinking acceptance, because many of us try to minimise the risk by driving carefully, and especially by buying safer cars with air bags and crumple zones. Indeed, with cars having a greater longevity than ever, buying a new car may have more to do with safety than reliability, unless you live next door to someone called Jones.

Yet we shrug off much greater risks. Otherwise why would so many of us smoke? A defining moment for many smokers can be surviving a heart attack, when the folly of our behaviour strikes home, and smoking is given up. Others take the more fatalistic view that the damage is already done, or they need a little pleasure in life. A new systematic review [1] tells us the fatalistic view just ain't so, and that smokers with coronary heart disease have an extra 1 in 10 chance of dying over five years because of their smoking.

Systematic review

The review sought studies with patients diagnosed with previous heart attack, or stable or unstable angina and who were smoking at baseline with smoking status well defined. Prospective cohort studies had to include current smokers at baseline, with smoking status measured to find who had quit smoking, in which the follow up was at least two years and with all cause mortality as an outcome measure.

The search strategy was extensive, examining nine electronic databases, and studies were not restricted by language.


There were 20 included studies with 12,600 patients, mostly using data collected in the 1960s and 1970s. Most cases were men (80%), and average cessation rate was 45%. Follow up ranged from two to 26 years, though most studies reported follow up of three to seven years, with a mean of five years.

Most studies involved follow up hospital case series, and reporting of smoking status was usually at some follow up appointment, though it was not validated, for example by biochemical measurement, in most studies. Most studies had a clear definition of the cardiac event. Loss to follow up was usually small. Size varied from under 100 to over 4,000 patients.

There were fewer deaths in quitters (18%) than in people who continued to smoke (27%), and the degree of reduction was consistent across all death rates reported (Figure 1). Results were broadly similar in all studies, and in six higher quality studies with about two-thirds of all patients (Table 1). Higher quality here was defined by a sample size of 500 smokers at baseline, with fewer than 15% dropouts, and with adequate or good control of confounding.

Figure 1: Death rates in patients with CHD who continued smoking, or who quit

Table 1: NNTs to prevent one death or reinfarction over five years by CHD patients quitting smoking compared with continuing to smoke

Number events/total (%)
Relative risk
(95% CI)
(95% CI)
Death (all studies)
0.64 (0.58 to 0.72)
12 (10 to 14)
Death (better studies)
0.71 (0.65 to 0.77)
15 (12 to 21)
Nonfatal reinfarction
0.68 (0.57 to 0.82)
28 (19 to 52)

A secondary outcome was nonfatal reinfarction, and while this was a less frequent event, the same degree of reduction was seen (Table 1).

Relative risk was reduced by about 30%, and the absolute risk by about 9% for mortality, translating into a number needed to treat of about 12 for one additional nonsmoking patient to be alive at five years.


This is a good review. It gives the consistent and unequivocal answer that smoking remains harmful after a coronary event. Carrying on smoking carries a five year risk of 1 in 10 of dying because you smoke. By contrast, the risk of dying on the roads over five years would be more like 1 in 4000. Smoking in those circumstances is 400 times more dangerous than driving.

Also interesting in the review is the discussion about limitations, and how limitations may affect the result. Most of them make the results more conservative. For instance, if people who said they had stopped smoking were lying, the quitter results would be worse than they should be, and the benefits of stopping smoking under-estimated.

There's another point here. The word quit is one with negative connotations, with dictionary definitions of releasing from obligation, or clearing off, as well as to leave off, which is what is meant in this context. A thesaurus gives alternatives like deliver, free, or liberate. Perhaps it is time to look for an alternative description of people who stop smoking, less clearing off and more deliverance, perhaps.


  1. JA Critchley, S Capewell. Mortality risk reduction associated with smoking cessation in patients with coronary heart disease. A systematic review. JAMA 2003 290: 86-97.

previous or next story