Transcript
Kardiovaskulær overdiagnostik Hálfdán Pétursson PhD Allmennlege Gøteborg
Cassius Coolidge. www.dogsplayingpoker.org
Blodtrykk Systolisk Bltr>200
Diastólískur BÞ>105 140/90
Samþætt áhættumat
1961: Risiko faktorer for CHD: - Røyke - Blodtrykk - Kolesterol Thiazid + Beta-blokker
Lancet 1964: Forebyggende Bltr behandling
1940
1950
1960
1970
1980
1990
2000
2010
2015
Lancet 1964 Feb 1;41:235-8.
Friske deltakere < 60 år Diastolisk blodtrykk >110 mmHg 22 menn og 49 kvinner
Blodtrykk Systolisk Bltr>200
Diastolisk Bltr>105
140/90
Risiko estimering
1961: Risiko faktorer for CHD: - Røyke - Blodtrykk - Kolesterol Thiazid + Beta-blokker
Prehypertension (120/80)
Lancet 1964: Forebyggende Bltr behandling
1940
1950
1960
1970
1980
1990
2000
2010
2015
Prospective Studies Collaboration. Lancet 2002
Risiko
Relative risiko kan fordubbles, men absolut risiko kan forts være veldig lav
Blodtrykk
Treatment benefit
More
Harm from treatment
Less Mild
Severe
Spectrum of Abnormality Figure reproduced from: Welch G, Schwartz L, Woloshin S. Overdiagnosis: Making people sick in the pursuit of health. Boston: Beacon; 2011
Systolic: 140-159 mmHg and/or Diastolic: 90-99 mmHg
Mild hypertension 2012
Ikke signifikant effekt på CVD eller mortalitet
Ann Intern Med 2015:162(3):184-91
OR:
Mortality: 0.78 (0.67-0.92) CVD events: 0.86 (0.74-1.01)
The Norwegian HUNT 2 Study >60 000 participants, 1995-97
100 90 80
BP≥120/80 mmHg
% of participants
70
BP≥130/80 mmHg
21%
60 50
Prehypertension
BP≥140/90 mmHg
Mild hypertension
BP>140/90 mmHg
40
40%
30
BP≥160/100 mmHg
20
Blood pressure lowering treatment
10 0 20
25
30
35
40
45
50
55
60
65
70
75 Age
Petursson H et al. J Eval Clin Pract 2009
- Bivirkninger - Labelling - Kostnad - Opportunity cost
Major Inclusion Criteria • At least 50 years old • Systolic blood pressure – – – –
SBP: 130 – 180 mm Hg on 0 or 1 medication SBP: 130 – 170 mm Hg on up to 2 medications SBP: 130 – 160 mm Hg on up to 3 medications SBP: 130 – 150 mm Hg on up to 4 medications
• Risk (one or more of the following) – Presence of clinical or subclinical CVD (not stroke) – Chronic Kidney Disease (CKD), defined as eGFR 20 – 59 ml/min/1.73m2 – Framingham Risk Score for 10-year CVD risk ≥ 15% – Not needed if eligible based on preexisting CVD or CKD
Major Exclusion Criteria • • • • • •
Stroke Diabetes Congestive heart failure (symptoms or EF < 35%) Proteinuria >1g/d CKD with eGFR < 20 mL/min/1.73m2 (MDRD) Adherence flags
IS THIS
?
• 1948 – Original cohort – 5 209 participants – Age 30-62
• 1948–2015 – 6 groups of participants – 3 generations – 15 447 participants
Inhabitants: 1950 – 28 086 2010 – 68 318
Observed risk (bold lines) with 95% confidence limits and predicted risk (broken lines) of incident non-fatal MI plus fatal coronary events.
Hense H et al. Eur Heart J 2003;24:937-945 The European Society of Cardiology
12 land 205 000 deltakere 2.7 millioner personår
Heterogene kohorter Rekruttering 1967-1991
Evne til å differensiere høy vs lav risiko individer = 70-80%
...Men hva med målgruppen?
Observert dødelighet lavere enn SCORE predikerer
Data fra HUNT 2 undersøkelsen N = 62 104 CVD diseases and number of risk factors
% 100
3.9%
0
90
1 risk factor
80
12.4%
70
2 risk factors
60
71.2%
Reality Hypertension guidelines 40GPs needed (BP only): GPs working (all diseases): ≥ 3 risk factors 30 99/100 000 adults 87/100 000 adults 12.5% 50
20
Petursson et al. BMC Family Practice 2009;10:70 CVD, diabetes and
10
treated hypertension
0 20
25
30
35
40
45
50
55
60
65
70
75
Age
Petursson et al. J Eval Clin Pract 2009;15:103-9
Cox regression Justerad for alder, røyk, blodtrykk - 52ooo hjertefriske N-Trønder - 0,5 million personår
Økning 1.0 mmol/L Risk reduction 6% P = 0.02 Økning 1.0 mmol/L HR: 0.98 95% CI: 0.93-1.03
Petursson H, et al. J Eval Clin Pract 2012:18:159-168
Petursson H, et al. J Eval Clin Pract 2012:18:159-168
Den magiske regressions hatten
Sant?
Riktigt?
Who cares?
Prevalence Men of statin use by gender and age during 2000-2010 Female 40
40
75-84 75-84
65-74 30
30
65-74 85+ 55-64 20
20
55-64
85+ 10
40-54
10
40-54
0
0 2000
2002 40-54
2004
2006 55-64
2008
2010 65-74
2000
2002
2004
2006
2008
2010
40-54 55-64 Wallach-Kildemoes et al. J Eval Clin Pract 65-74 2015
Sakshaug S (ed). Drug Consumption in Norway 2010-2014. Oslo: Norwegian Institute of Public Health.
21 000 persons Total cholesterol HR Stroke: 0.78 (0.69-0.89) per 1 SD increase IHD: 1.03 (0.92-1.14)
• Oppmerksomme begrensninger – Individuelle leger – Modelling studies
• Forbedre kalkulatorene – Lokalt • Kalibrering
– Globalt • Ikke-lineære sammenheng • Flere/nye prediksjons faktorer
OST (Osteoporosis Self-assessment Tool) DEXA if: Weight (kg) – Age < 10 (♂ < 20)
P4 Medicine Predictive, Preventive, Personalized and Participatory
‘If you talk to technology gurus in California and ask what’s going to change in the next two decades, they say “in 20 years’ time, no doctor will ever give a diagnosis”. ’
“You can get three hundred thousand biomarkers from a single drop of blood, so why would you depend on a human brain to calculate what that means when a computer can do it for you?”
‘I think it’s really important that we’re ready in the NHS to harness the power of data to give us more accurate diagnoses, in particular with that example.’
’What this will mean, is we can identify problems before they’re symptomatic and therefore have a much better chance of tackling them. So it’s a pretty exciting prospect but there’s lots of work to do.’
• 128 hypotetiske pasienter • 25 risiko kalkulatorer • 3 risiko kategorier Gjennomsnittlig concordance: 67% Pasienter som havner i alle 3 kategorier:
41%
Norske retningslinjer 2009 HUNT 3 (2006-08)
100 %
Recommendations
90 %
Regular
80 % 70 %
Follow-up
58.6%
Treatment
60 % 50 %
Established treatment
8.7%
40 %
8.2%
30 %
BP drugs Diabetes
24.5%
20 %
CVD
10 % 0% 40
45
50
55
60
65
Age
Figure 3. Established treatment and treatment recommandations by age, taking additional risk factors, not included in the NORRISK chart, into account. Petursson et al. Unpublished data
IHD mortality by age
N = 900 000 12 million personyears
Stroke mortality by age
N = 900 000 12 million personyears
Stroke mortality by SBP
Kolesterol og CVD dødelighet: ? Kolesterol og total dødelighet: ?
N = 900 000 12 million personyears
