Statins in primary prevention and all-cause mortality

Journal reference: Ray KK, Seshasai SR, Erqou S, et al.  Statins and all-cause mortality in high-risk primary prevention: a meta-analysis of 11 randomized controlled trials involving 65 229 participants. Archives of Internal Medicine, 2010; 170(12): 1024-31 ^[1]

Link: http://archinte.ama-assn.org/cgi/content/abstract/170/12/1024

Published: 28 June 2010

Evidence cookie says...

Statins were not associated with a benefit to all-cause mortality in a primary prevention setting. Although there is some ambiguity with the data, it is very unlikely that statins are of clinical benefit in patients with low and moderate absolute cardiovascular risk. Until more definitive evidence becomes available, statins in the primary prevention setting should be reserved for those patients who are at high risk according to the 2005 lipid management guidelines [2]

Article details:

Study design:

meta-analysis of randomised controlled trials

Study aim:

to reliably determine if statin therapy reduces all-cause mortality among intermediate to high-risk individuals without a history of cardiovascular disease

Study conclusion:

This literature-based meta-analysis did not find evidence for the benefit of statin therapy on all-cause mortality in a high-risk primary prevention set-up ^[1].

Critical appraisal:

Are the trial results valid?

Internal validity: Wikipedia

Is it unlikely that important, relevant studies were missed?

Yes.

• two major databases were searched (MEDLINE and the Cochrane Collaboration)
• bibliographies of retrieved studies were searched
• authors of further trials that included participants with cardiovascular disease at baseline, were asked for provide further information; to see if unpublished tabular information on the subset of participants without baseline cardiovascular disease was available

Were the criteria used to select articles for inclusion appropriate?

Yes.

Inclusion criteria were:

• randomised trials of statins vs placebo/control
• trials that collected information on all-cause mortality; and
• trials conducted among individuals without cardiovascular disease at baseline

Were the included studies sufficiently valid for the type of question asked?

Yes.

• the studies were all large randomised trials of statins vs placebo/control

Were the results similar from study to study?

Yes.

• there was only limited evidence of study heterogeneity in the correlation of statins on all-cause mortality
• I²=23%; (95% CI 0-61%), P = 0.23
• interpretation: only a mild (none to modest if considering the confidence interval) proportion of the variation of results is due to differences between the included studies; i.e., the studies were similar enough that it is valid to group the data together for analysis

What were the results?

Primary outcome

All-cause mortality:

• placebo/control group: 11.4 per 1000 person-years
• statin group: 10.7 per 1000 person-years
• Risk ratio (RR) for all-cause mortality associated with use of statins:
  • RR = 0.91 (95% CI 0.83 – 1.01)
  • interpretation: statin use was associated with a small relative risk reduction of all-cause mortality but this was not statistically significant
• Number needed to treat (NNT) (note: calculated by Morsels of Evidence):
  • treated with statin compared to placebo for 5 years to prevent one additional death:
  • NNT = 286 (95% CI 261 to infinity)
  • note: it should be reminded that the RR was not statistically significant so the 95% confidence interval for the real NNT includes infinity.

Other outcomes:

Relationship between lipid levels and all-cause mortality:

• no significant relationship was observed between the mean baseline levels of LDL-C and the relative reduction in all-cause mortality across studies (P = 0.97)

Will the results help me care for my patient?

External validity: Wikipedia

Are the participants different to my patient?

Study population:

• 65 229 subjects in predominantly Western populations
• no baseline cardiovascular disease
• some participants with type 2 diabetes mellitus
• mean age ranged from 51 to 75 years
• average baseline LDL-C was 3.6 mmol/L (138 mg/dL)
• relatively short mean follow up period, 3.7 years
• placebo/control group all-cause mortality rate:
  • 11.4 per 1000 person-years
  • 5.7% per 5-years
• it is not clear what the average absolute cardiovascular risk category these patients were in as it wasn’t a measured outcome
  • however, using the assumption that, (i) the majority of deaths are secondary to cardiovascular disease, and (ii) that at least half of those with cardiovascular disease would not have died, it can be inferred that that these patients had a moderate to high absolute 5-year risk of a cardiovascular event (> 10%)

Is the treatment feasible?

Yes.

• statin therapy in primary prevention is a potentially feasible treatment
• they are widely used for the treatment and prevention of cardiovascular disease both among individuals with established disease, and among high-risk healthy individuals
• it should be noted that in most situations, statins cannot be prescribed under the pharmaceutical benefits scheme (PBS) for primary prevention

Were all the clinically important outcomes considered?

No.

• the study does not measure the effects of statins on non-fatal cardiovascular outcomes

Are the treatment benefits worth the potential harms/costs?

Unclear, probably no.

• the benefits of statin therapy for all cause mortality in the primary prevention setting is possibly more modest than previously thought
• the population appears to be at high risk of death (5.7% risk over 5-years), but their absolutely cardiovascular risk can only be inferred (likely high)
• this study demonstrated a small, non-statistically significant benefit  to all-cause mortality
• even using results in the most favourable end of the confidence interval, the NNT with statin over placebo for 5-years is substantial (over 260)
• although no cost-effectiveness analysis was performed, it is very unlikely that statin therapy in primary prevention is cost-effective overall

Study weaknesses (summary)

• relatively short mean follow-up
• the study did not evaluate the effects of statins on cardiovascular mortality or non-fatal cardiovascular disease
• the study does not report the absolute cardiovascular risk of the study participants, and nor are there any results that divide the participants into groups according to baseline absolute cardiovascular risk
  • this would have improved the direct applicability of the results
• the lack of participant level data precludes any useful subgroup analyses
• authors attempted but were unable to obtain information from 4 published studies fulfilling the study criteria that would have contributed further information on approximately 3700 persons

Study strengths (summary)

• no material heterogeneity across the studies
• detailed protocol developed a priori, with thorough search of published and unpublished
• large number of participants
• authors obtained unpublished data from the authors of several original articles

Biases and conflicts of interests

• Ray KK (primary author) received honoraria for lectures and advisory boards from the majority of companies that market lipid-lowering agents
• various other authors received grants, fees, research support and honoraria from various pharmaceutical companies

Clinical relevance to primary health care

Statins are widely used for the treatment and prevention of cardiovascular disease, in individuals with established disease (secondary prevention), and in the primary prevention setting in individuals at an elevated risk of a cardiovascular event.

This study did not find statin therapy to be beneficial on all-cause mortality in a primary prevention setting, in a population without established cardiovascular disease but at an elevated risk of mortality (5.7% risk per 5-years). It is probable that any mortality benefits are likely to be more modest than previously perceived.

What do these results mean for Australian general practitioners?

• The grouped nature of these results, that the participants were not separated into 5-year absolute cardiovascular risk groups at baseline, and the necessary inferences on overall cardiovascular risk from the all-cause mortality rate; leads to increasing ambiguity in the validity of the findings when applied to patients at the highest cardiovascular risk.
• This study does not provide definitive evidence that overturns the current recommendation that individuals free from existing cardiovascular disease who nonetheless are at high risk of an event (i.e., 5-year absolute risk ≥ 15%, or, absolute risk 10-15% in family history of premature coronary heart disease or the metabolic syndrome) should have lipid-modifying therapy ^[2].

However, this study does support the notion that primary prevention of cardiovascular disease with statins in populations at moderate or low risk is of little or no benefit. Even if the number needed to treat (NNT) at the most favourable extent of the confidence interval is true, the average Australian GP will need to treat their entire patient base who are older than 50, who do not currently have established cardiovascular disease or other indications, with statins over 5-years to prevent a single death.

In the primary prevention setting, statins should be reserved for those patients who are at high risk according to the 2005 lipid management guidelines ^[2].

References

1. Ray KK, Seshasai SR, Erqou S, et al.  Statins and all-cause mortality in high-risk primary prevention: a meta-analysis of 11 randomized controlled trials involving 65 229 participants. Archives of Internal Medicine, 2010; 170(12): 1024-31
2. National Heart Foundation of Australia and the Cardiac Society of Australian and New Zealand. Position statement on lipid management – 2005. Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society of Australia and New Zealand, Elsevier Inc. 2005

Co-author: Michael Tam