Grossi postulates that plausibility might be a concept better understood by patients than probability.(1) Can it be conveyed to a patient whether it is plausible that they would benefit from taking a drug?
The number needed to treat (NNT) is a well established and easily understood means of describing the effectiveness of a treatment.(2,3) Based on the change in absolute risk, it represents the probability of successfully treating a patient. However, many treatments are for chronic disease and thus continuous and lifelong. NNTs are fine for one off treatments such as a surgical procedure or a course of antibiotics but begin to lose meaning when there is a dimension of time as well. Laupacis and colleagues suggested dealing with this by using a standard duration such as 5 years (‘NNT:5’).(2)
Many epidemiological studies, such as those investigating the effects of hormone replacement therapy, use person years as denominator to capture the time dimension. An effectiveness measure based on this would be ‘person years needed to treat’, the PYNT or ‘NNT:1’. For the professional this has equal meaning, be it 100 people for one year or 1 person for 100 years. For the individual the latter has more meaning. Indeed, when discussing with an individual in order to obtain informed consent it might be helpful to the patient to be told the number of years they would need to take a drug to benefit. In the case of a statin at the current NICE threshold for primary prevention (20% CVD event risk over 10 years) the PYNT would be about 150 years.(4) Average risk in the ‘4S’ study of people with existing coronary heart disease was ~50% over 10 years, but even if we assume a 10 year risk of 100%, the PYNT is ~30, 3 times the actuarial life expectancy (ALE) of an average 75 year old English male.(5,6)
For a 59 year old male (me) at the 20% threshold there are 6.9 ALEs to a 150 sized PYNT.(6) That’s nearly 7 of my lifetimes: is it plausible that I would benefit from a statin before I died? I think I’ll stick to a glass of red wine each day, though not a pint glass.
References. 1 Grossi E: Medical concepts related to individual risk are better explained with "plausibility" rather than "probability". BMC Cardiovasc Disord 2005; 5:31. 2 Laupacis A, Sackett, DL, Roberts RS. An Assessment of Clinically Useful Measures of the Consequences of Treatment. N Engl J Med 1988; 318:1728-1733 3 Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence Based Medicine. Churchill Livingstone. London 1997. 4 Lipid modification: Cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease. NICE 2010. CG67. 5 Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-9. 6 Office for National Statistics. http://www.statistics.gov.uk/downloads/theme_population/Interim_Life/ILTEng0608Reg.xls
Richard Richards, Assistant Director of Public Health, Derbyshire, England; richard.richards@derbyshirecountypct.nhs.uk
How much ALE to the PYNT?
12 August 2010
Grossi postulates that plausibility might be a concept better understood by patients than probability.(1) Can it be conveyed to a patient whether it is plausible that they would benefit from taking a drug?
The number needed to treat (NNT) is a well established and easily understood means of describing the effectiveness of a treatment.(2,3) Based on the change in absolute risk, it represents the probability of successfully treating a patient. However, many treatments are for chronic disease and thus continuous and lifelong. NNTs are fine for one off treatments such as a surgical procedure or a course of antibiotics but begin to lose meaning when there is a dimension of time as well. Laupacis and colleagues suggested dealing with this by using a standard duration such as 5 years (‘NNT:5’).(2)
Many epidemiological studies, such as those investigating the effects of hormone replacement therapy, use person years as denominator to capture the time dimension. An effectiveness measure based on this would be ‘person years needed to treat’, the PYNT or ‘NNT:1’. For the professional this has equal meaning, be it 100 people for one year or 1 person for 100 years. For the individual the latter has more meaning. Indeed, when discussing with an individual in order to obtain informed consent it might be helpful to the patient to be told the number of years they would need to take a drug to benefit. In the case of a statin at the current NICE threshold for primary prevention (20% CVD event risk over 10 years) the PYNT would be about 150 years.(4) Average risk in the ‘4S’ study of people with existing coronary heart disease was ~50% over 10 years, but even if we assume a 10 year risk of 100%, the PYNT is ~30, 3 times the actuarial life expectancy (ALE) of an average 75 year old English male.(5,6)
For a 59 year old male (me) at the 20% threshold there are 6.9 ALEs to a 150 sized PYNT.(6) That’s nearly 7 of my lifetimes: is it plausible that I would benefit from a statin before I died? I think I’ll stick to a glass of red wine each day, though not a pint glass.
References.
1 Grossi E: Medical concepts related to individual risk are better explained with "plausibility" rather than "probability". BMC Cardiovasc Disord 2005; 5:31.
2 Laupacis A, Sackett, DL, Roberts RS. An Assessment of Clinically Useful Measures of the Consequences of Treatment. N Engl J Med 1988; 318:1728-1733
3 Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence Based Medicine. Churchill Livingstone. London 1997.
4 Lipid modification: Cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease. NICE 2010. CG67.
5 Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-9.
6 Office for National Statistics. http://www.statistics.gov.uk/downloads/theme_population/Interim_Life/ILTEng0608Reg.xls
Richard Richards, Assistant Director of Public Health, Derbyshire, England; richard.richards@derbyshirecountypct.nhs.uk
Competing interests
None declared