Download presentation
Presentation is loading. Please wait.
Published byDoreen Harrison Modified over 9 years ago
1
Changes in lifestyle risk factors: Health and economic impact as estimated by the population based RHS- model Inna Feldman, PhD nna.feldman@kbh.uu.se Social Pediatrics Women’s and Children’s Health Uppsala University, Sweden
2
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.
3
Estimation of future costs Health (risk factors) Morbidity Costs Now… In future…
4
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
5
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 %. * 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.
6
Time horizon – 10 yeas, from year 5. Relative risks (RR) are changing linearly, from RR to 1 during 10 years; RR i = (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), p2 i =(p2/10)* i The total reduction of incidence: Where - reduction in incidence during the year i. Time horizon
7
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 QALY 1 & DALY 2 – health economics measures RHS - model 1 Swedish and international studies, 2 Salomon et al, 2012
8
The model diseases Obesity, BMI>30 Daily smoking Physical inactivity Risky consu mption of alcohol ICD-10 code Diabetes type 2 xxx E11 Ischemic heart disease xxx I20, I24, I25 Stroke xxx I61, I63, I64 COPD xx J40-J44 Depression xxxx F32-F33 Hip fracture xxx 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 xxxx C18 Lung x C34 Breast xxx x C50 Prostate xx C61 Esophageal x C15 Liver x C22 Annual incidence in the diseases in the Swedish population.
9
Relative risks – some examples Relative risks in for daily tobacco smoking MenWomenSources 20-4445-6465-84 20-4445-6465-84 Diabetes type 2 1.2 Willi et al, 2007 Ischaemic hd3.11.81.43.62.11.5Prochaska & Hilton, 2012 Stroke2.81.91.53.22.11.3 Colditz et al, 1998; Robbins et al, 1994 COPD10.612.311.89.310.87.5Lindberg et al, 2006 Depression1.1 Buden et al, 2010 Hip fracture1.8 Marks, 2010 Cancers: Colon1.2 Giovannucci, 2001; Parkin, 2011 Lung26.428.021.616.114.110.6Parkin, 2011 Breast---1.1 Terry et al, 2002 Prostate1.1 - - -Huncharek et al, 2010
10
Costs: average annual costs Health care: Swedish national and regional registers: inpatient, specialist outpatient, and primary health care Municipal care: estimated based onMunicipal 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,
11
QALY och DALY weights, for a year spent in disease QALY weightDALY weight Diabetes type 2 0.660.03 Ischaemic heart disease 0.600.06 Stroke 0.520.08 COPD 0.730.19 Depression 0.680.41 Hip fracture 0.670.31 Liver cirrhosis 0.620.19 Epilepsy 0.640.32 Mental and behavioural disorders due to use of alcohol 0.700.39 Cancers: 0.29 Colon 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
12
Outcomes Health gains : –decreased incidence –increased QALYs –increased DALYs Change in societal costs: –health care –municipality care –sickness insurance
13
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
14
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
15
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
16
RHS-model as a computer application!
17
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
18
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
19
Development plans To include the population prognosis function More risk factors? Web-based application, PC and Mac ???
20
P.S. Similar models: Dutch RIVM model (Feenstra et al, 2011) Australia (Cadilhac et al, 2011)
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.