Changes in lifestyle risk factors: Health and economic impact as estimated by the population based RHS- model
Inna Feldman, PhD
[email protected] Pediatrics Women’s and Children’s Health Uppsala University, Sweden
Problem to address
• Is it possible to estimate societal cost savings given a change in population lifestyles?
Support for health promotion specialists and decision-makers.
Risks, Health and Societal costs – RHS model*• Simulates changes in incidence and
related societal costs of several chronic diseases following changes in prevalence:
BMI>30, obesity
Daily tobacco smoking
Lack of physical activity, less than 2h/week
Risky consumption of alcohol (AUDIT)
* Feldman I. and Johansson P. The Swedish RHS-model (Risk factors, health and societal costs). Technical report. Available at http://www.hfsnatverket.se/lib/get/doc.php?id=15399bcf46ed95
Calculation methods• Based on Relative risks (RR) and Potential Impact Fractions (PIF) *
RR= P exposed / P non-exposed
RR men, age 50-64( smoker, stroke)=2.6
• The incidence rate of the disease after the change in
the related risk factor (I*):
Daily smoking and stroke:
P=0.13 (13%); P*=0.1 (10%); RR=2,6 IF=0,04
A reduction in prevalence of daily smoking from 13 % to
10 % results in a reduction in the incidence of stroke by 4 %.
( * )(RR 1)
( 1) 1
P PPIF
P RR
* (1 )I I PIF
* Morgenstern H, Bursic ES. A method for using epidemiologic data to estimate the potential impact of an intervention on the health status of a target population. J Community Health. 1982; 7:292-309.
• Time horizon – 10 yeas, from year 5.• Relative risks (RR) are changing linearly, from RR to 1 during
10 years; RRi = (RR/10)* i, where i - number of the year, i= 5 to 10
• Risk factors prevalence is changing linearly, from the year five (i=5 to 10), p2i=(p2/10)* i
The total reduction of incidence:
Where - reduction in incidence during the year i.
Time horizon
5( ) ( )
n
iF n f i
Risk factors: BMI>30, obesity Daily tobacco smoking Lack of physical activity, less than 2h/week Risky consumption of alcohol (AUDIT)
Source: Swedish population survey
Age groups: adults, 20-84 years old (3 age groups), men and women
Base for economic consequences:: Lower number of new cases (reduced incidence) due to positive development of risk factors
QALY1 & DALY2 – health economics measures
RHS - model
1 Swedish and international studies, 2 Salomon et al, 2012
The model diseases Obesity,
BMI>30Daily
smokingPhysical inactivity
Risky consumption
of alcoho
l
ICD-10 code
Diabetes type 2 x x x E11
Ischemic heart disease x x x I20, I24, I25
Stroke x x x I61, I63, I64
COPD x x J40-J44
Depression x x x x F32-F33
Hip fracture x x x S72.0-S72.2
Liver cirrhosis x K70, K74
Epilepsy x G40- G41
Mental and behavioral disorders due to use of alcohol
x F 10
Cancers:
Colon x x x x C18
Lung x C34
Breast x x x x C50
Prostate x x C61
Esophageal x C15
Liver x C22
Annual incidence in the diseases in the Swedish population.
Relative risks – some examples
Relative risks in for daily tobacco smokingMen Women Sources
20-44 45-64 65-84 20-44 45-64 65-84
Diabetes type 2
1.2 1.2 1.2 1.2 1.2 1.2 Willi et al, 2007
Ischaemic hd 3.1 1.8 1.4 3.6 2.1 1.5 Prochaska & Hilton, 2012
Stroke 2.8 1.9 1.5 3.2 2.1 1.3Colditz et al, 1998; Robbins et al, 1994
COPD 10.6 12.3 11.8 9.3 10.8 7.5 Lindberg et al, 2006
Depression 1.1 1.1 1.1 1.1 1.1 1.1 Buden et al, 2010
Hip fracture 1.8 1.8 1.8 1.8 1.8 1.8 Marks, 2010
Cancers:
Colon 1.2 1.2 1.2 1.2 1.2 1.2Giovannucci, 2001; Parkin, 2011
Lung 26.4 28.0 21.6 16.1 14.1 10.6 Parkin, 2011
Breast - - - 1.1 1.1 1.1 Terry et al, 2002
Prostate 1.1 1.1 1.1 - - - Huncharek et al, 2010
Costs: average annual costs• Health care: Swedish national and regional
registers: inpatient, specialist outpatient, and primary health care
• Municipal care: estimated based on level of dependency 1,2
• Sickness insurance: estimated based on the level of absence due to sickness 2, 80% of lost income, based on 24 000 SEK
Costs
1) Lindholm et al, 2012, 2) Salomon et al, 2012
Reflect costs for three Swedish sectors: the regional healthcare, the local authorities and the national social insurance,
QALY och DALY weights, for a year spent in disease
QALY weight DALY weight
Diabetes type 2 0.66 0.03Ischaemic heart disease 0.60 0.06Stroke 0.52 0.08COPD 0.73 0.19Depression 0.68 0.41Hip fracture 0.67 0.31Liver cirrhosis 0.62 0.19Epilepsy 0.64 0.32Mental and behavioural disorders due to use of alcohol 0.70 0.39Cancers: 0.29Colon 0.67 Lung 0.56 Breast 0.76 Prostata 0.69 Oesophageal 0.82 Liver 0.82
Sullivan et al, 2011, web table 3; Salomon et al, 2012, table 2
Outcomes
• Health gains :– decreased incidence– increased QALYs– increased DALYs
• Change in societal costs:– health care – municipality care – sickness insurance
Summary of model input and output data
The fixed parameters are:• Relative risks for the 15 diseases, subject to the
risk factor prevalence, for the six gender-specific age groups
• Incidence in the 15 diseases, for the six gender-specific age groups
• Annual societal costs for a person with a certain disease
• Annual health effects, in QALYs and DALYs, for a person with a certain disease
Summary of model input and output data
The input parameters are:• Number of population, for the six gender-
specific age groups• Current prevalence of the four risk factors in the
six gender-specific age groups, expressed in percent
• Desired prevalence of the four risk factors in the six gender-specific age groups, expressed in percent
Summary of model input and output data
The model outputs for the 5 year horizon:• Changes in number of incident cases, in year 5• Changes in societal costs, total as well as per sector,
in year 5• Changes in health effects, in QALYs and DALYs, in
year 5
The model outputs for the n-year horizon (n=6 to 10):• Changes in number of incident cases, accumulated
from year 5 to year n• Changes in societal costs, total as well as per sector,
accumulated from year 5 to year n• Changes in health effects, in QALYs and DALYs,
accumulated from year 5 to year n
Strengths
• Can include as many diagnoses as we have data for:– Incidence– Risk factors and RR– Costs
• Easy to understand and to use, can be applied to local data
Limitations
• Based on the population at baseline, should include population prognosis
• Time aspect, more careful estimation
• Some risk factors significantly correlate, overestimation
• The model estimates only reduction in morbidity incidence, changes in life style affect morbidity prevalence, underestimation
Conclusions
• The decrease in the prevalence of risk factors can result in cost savings for the society
This model can be adapted to different
populations by taking into account the existing
age structure and the prevalence of risk factors
The model can be extended/adapted for
different diagnoses and risk factors
Development plans
• To include the population prognosis function
• More risk factors?• Web-based application, PC and Mac• ???