|Cost effectiveness and equity of a community based cardiovascular disease prevention programme in Norsjo, Sweden
|Lindholm L, Rosen M, Weinehall L, Asplund K
This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.
Community based cardiovascular disease prevention (educational) programme.
Type of intervention
Screening; Primary prevention; Secondary prevention.
Economic study type
The study population was taken from a rural community in Northern Sweden with an 80% excess mortality rate in cardiovascular disease compared with the Swedish national average and crude prevalences of hypercholesterolaemia (>= 6.6 mmol
Community. The economic study was carried out in Sweden.
Dates to which data relate
Effectiveness data were collected over the period 1985-1990. Resource use data/costs were derived from 1987-1994 data.1992 prices were used
Source of effectiveness data
Effectiveness data were derived from a single study.
Link between effectiveness and cost data
Costing was not undertaken on the same patient sample as that used in the study.
All individuals in Norsjo at the age of 31, 41, 51 and 61 were invited to be screened in each year.5% of individuals refused to participate and, in total, 1,498 individuals were screened. For the control population a stratified sample of 2,000 individuals was randomly selected from population records in 1986 and 1990 of individuals in the WHO MONICA project. In 1986, 1,625 participated whilst 18.7% refused and in 1990, 1,583 participated, 20.8% refusing. The use of power calculations was not reported.
This was a before-and-after study. The duration of follow up of the treatment cohort was to 1995 or death. No loss to follow up was stated.
Analysis of effectiveness
The basis of the analysis of effectiveness was not stated. The primary health outcomes were risk factor trends in CHD, serum cholesterol levels and blood pressure.
In the study population mean serum cholesterol levels fell by 1.0 mmol/l in men (p=0.01) and 1.4 mmol/l (p=0.03) in women. In the control group no change took place in men and, in women, levels fell by 0.2 mmol/l, (p=0.01). In all groups other than men aged 50 and women aged 40, predicted cumulative incidence of CHD fell. Specifically for men aged 30 this fell by 0.0032 (p=0.037), for men aged 40 by 0.0297 (p=0.0001), and for men aged 60 by 0.04866 (p=0.0005). For women aged 30 CHD incidence fell by 0.0006 (p=0.0116), for those aged 50 by 0.0067 (p=0.0090) and those aged 60 by 0.0180 (p=0.0017). This decrease in incidence translated to 4.5 cases of cardiovascular disease in Norsjo in any year for the population between 25 and 64. The proportion of those with low serum cholesterol levels increased significantly to 6.5% (p=0.023) and estimated excess mortality caused by low cholesterol in Norsjo was 0.02 cases for both genders.
The change in cholesterol levels is due to the intervention as there was little change in the general population. Cardiovascular community intervention programmes are beneficial to populations and outweigh the negative effects of reduced cholesterol levels. All social groups benefit to the same extent.
Multi-variate logistic regression models were used to estimate the 8-year (1980-1998) cumulative incidence of coronary heart disease (CHD) and associated costs. The model was used to differentiate effects across socioeconomic groups and to predict results for the 1985-1990 period, followed by scenario analyses based on assumptions on the course of effects and benefits for the period 1990-1998.
Measure of benefits used in the economic analysis
The measure of benefits was life years saved, calculated by use of a regression model. Life years saved were estimated until 1999 (starting in 1985).
The costs of the screening programme and life style advice were calculated alongside the savings associated with events prevented. Specifically costs related to screening:labour and laboratory costs. Other costs measured were those of the dietician, marketing of groceries, authority, and screening circles. Costs for screening were based on time measurements. Staff cost time was based on employer's costs. Screening circle costs were derived from information from the responsible organisations and the annual cost relates to five circles of 30 hours each with ten people. Marketing and authority costs were taken from budget analysis and interviews and inquiries with relevant personnel. The costs of treatments for myocardial infarction were based on Swedish randomised control trials regarding drug therapy published into 1987 and 1992. Angina pectoris costs were derived from cost effectiveness analyses in part from the Swedish Council on Technology Assessment in Health Care.1992 prices and a 5% discount rate were used. Costs were derived from a model which covered the period 1985-1999.
Costs were discounted. The time costs of programme participants was derived from the employer's costs of an industrial worker (productivity losses). Similarly, lost (gained) production from avoidance of cardiovascular events was also included in the analysis. Total costs were derived from a model covering the period 1985-1999. Quantities were not reported separately, although the authors referred to a previously published study. The price year was 1992.
UK pounds sterling (). The conversion rate used was SEK12.00 to 1.00.
A best-worst scenario and varying assumptions of benefit/effect duration were investigated. The benefits were reported both discounted and undiscounted.
Estimated benefits used in the economic analysis
In the 'plateau' version of the programme (where cholesterol reduction is assumed to end in 1990) 30 life years were saved over the period 1985 - 1995. This would prevent 4.5 cardiovascular disease cases in any one year.
The annual costs of the programme based on 250 screenings per year was 51,050. The ten year cost of the programme after discounting was 3,000.
Synthesis of costs and benefits
Under the 'plateau' case (see 'Estimated benefits...' above), the discounted cost per discounted life year saved by the 10-year programme was 1,200 (850, without discounting benefits). Similarly, the worst case scenario, with 50% higher costs and 50% lower avoided treatments, resulted in a cost-per-life-year-saved figure of 13,950 (9,500 with no discounting of benefits).
The programme is beneficial to the community, in particular to the upper middle aged groups. In the 55-64 age range 8-9 cardiovascular events per 1000 men are estimated to have been avoided each year in Norsjo at an acceptable cost.
The effectiveness study may be subject to bias in the estimates of benefits, due to the non-random and retrospective nature of the study design. It is open to question whether participants' time in the study should be valued in terms of the costs to an employer of an industrial worker as not all participants in the study would be employed in this category. The authors do not appear to have carried out an incremental cost-effectiveness analysis which could be considered essential to determine the cost-effectiveness of extending the screening programme to ever increasing sections of the population with different levels of risk.
Implications of the study
Community cardiovascular disease programmes may be effective, however further studies are required to determine which, if any, aspect of the programme is significantly more effective.
Lindholm L, Rosen M, Weinehall L, Asplund K. Cost effectiveness and equity of a community based cardiovascular disease prevention programme in Norsjo, Sweden. Journal of Epidemiology and Community Health 1996; 50: 190-195
Subject indexing assigned by NLM
Adult; Age Factors; Cardiovascular Diseases /economics /mortality /prevention & Cholesterol /blood; Cost-Benefit Analysis; Cross-Sectional Studies; Female; Humans; Logistic Models; Male; Middle Aged; Primary Prevention /economics; Prospective Studies; Risk Factors; Sex Factors; Sweden; control
Date bibliographic record published
Date abstract record published