|
General practitioners' role in preventive medicine: scenario analysis using alcohol as a case study |
Doran C M, Shakeshaft A P, Fawcett J E |
|
|
Record Status 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. Health technology The study examined the role of general practitioners (GPs) in the detection and intervention of at-risk drinking among patients. At-risk drinking was defined as a score of 5+ for males and 4+ for females to the first three items of the Alcohol Use Disorders Identification Test.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised a hypothetical cohort of patients presenting at a GP for any reason. Implicitly, patients were at least 14 years of age.
Setting The setting was primary care. The economic study was conducted in Australia.
Dates to which data relate The effectiveness and resource use data were gathered from literature published between 1994 and 2002. The cost data were taken from a national source published in 2001 but relating to 2000. The price year was 2000.
Source of effectiveness data The estimates for the final outcomes were derived from a synthesis of published studies and authors' assumptions.
Modelling A decision analysis model was used to estimate the costs and benefits of each strategy. The structure of the tree was presented. The events nodes were as follows. First, the at-risk drinker visited a GP. Second, the GP established the patient as being at risk. Third, the GP intervened with detected at-risk drinkers. Finally, the at-risk drinker changed behaviour to reduce risk level. The time horizon of the analysis was established so as to cover costs for a single consultation. No further details of the model were provided.
Outcomes assessed in the review The parameters obtained from published sources and used in the model were:
the proportion of at-risk drinkers visiting a GP in the last 12 months;
the proportion of at-risk drinkers detected by a GP;
the proportion of at-risk drinkers offered an intervention by a GP; and
the proportion of at-risk drinkers reducing their drinking.
Study designs and other criteria for inclusion in the review The authors did not describe how they identified the sources that provided the values of the model parameters. No specific criteria for inclusion in the review were stated. The review included two reports, statistical information, one study of unspecified design, and guidelines based on a summary of research.
Sources searched to identify primary studies Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included The values for the parameters in the model were obtained from 5 published sources.
Methods of combining primary studies A narrative explanation described how the study results were combined to estimate the proportion of at-risk drinkers visiting a GP. None of the other parameters were combined as they were derived from a single source.
Investigation of differences between primary studies No differences between the primary studies were investigated.
Results of the review The following probabilities were used in the model:
the proportion of at-risk drinkers visiting a GP in the last 12 months was 81% for males and 89% for females;
the proportion of at-risk drinkers detected by a GP was 43% for males and 29% for females;
the proportion of at-risk drinkers offered an intervention by a GP was 21% for males and 11% for females; and
the proportion of at-risk drinkers reducing their drinking was 45% for males and 45% for females who received a GP intervention, and, 24.1% for males and 20.3% for females who were at-risk drinkers not detected by a GP.
Methods used to derive estimates of effectiveness The authors made an assumption to obtain one of the estimates of effectiveness.
Estimates of effectiveness and key assumptions The model assumed that 34.6% of males and 32.7% of females whose at-risk drinking was detected, but who did not receive a GP intervention, reduced their drinking anyway. These percentages were selected as they represent the mid-point between the estimates of those patients reducing their alcohol consumption after receiving a GP intervention and those who reduced their drinking even though their at-risk behaviour was undetected.
Measure of benefits used in the economic analysis The measure of health benefit used was the number of at-risk drinkers that reduced their alcohol consumption after detection and intervention by a GP. The number of at-risk drinkers detected and the number offered an intervention were also reported. These outcomes were obtained directly from the model.
Direct costs The authors considered the direct costs to the health service. The only cost included in the analysis was the cost of the GP visit. The basic fee for a straightforward consultation was subtracted from the fees for longer consultations to reflect the GPs time costs for detection, and for detection and intervention. The costs were obtained from the Medical Benefits Schedule, and were reported separately from other model parameters. Discounting was not applied, which was appropriate given that a short time horizon appears to have been used. The costs were expressed in 2000 prices. The direct costs estimated were the total costs of each strategy.
Statistical analysis of costs No statistical analysis of the costs was conducted.
Indirect Costs The indirect costs were not included.
Currency Australian dollars (Aus$).
Sensitivity analysis The alternative levels of detection, intervention and effectiveness in scenarios 1, 2 and 3 were, in effect, a one-way sensitivity analysis. Likewise, the combinations in scenario 4 were, effectively, two-way sensitivity analyses. These analyses investigated the model structure. The authors did not justify the ranges over which the variables were tested.
Estimated benefits used in the economic analysis Of the 3,213,435 at-risk drinkers visiting a GP, 1,189,050 were detected as being at risk by a GP and 209,778 detected drinkers were then offered an intervention.
At the current level, 94,400 at-risk drinkers reduced their alcohol consumption after detection and intervention by a GP.
In scenario 1, 5%, 10% and 100% increases in detection resulted in 106,486, 118,572 and 241,722 at-risk drinkers, respectively, reducing their alcohol consumption after detection and intervention by a GP.
In scenario 2, 5%, 10% and 100% increases in intervention resulted in 121,154, 147,908 and 535,072 at-risk drinkers, respectively, reducing their alcohol consumption after detection and intervention by a GP.
In scenario 3, 5%, 10% and 100% increases in effectiveness resulted in 104,889, 115,378 and 209,778 at-risk drinkers, respectively, reducing their alcohol consumption after detection and intervention by a GP.
In scenario 4, a 5% increase in detection and intervention resulted in 136,855 at-risk drinkers reducing their alcohol consumption after detection and intervention by a GP. Similarly, a 5% increase in intervention and effectiveness and a 5% increase in detection and effectiveness resulted in 134, 615 and 118,318 at-risk drinkers, respectively, reducing their alcohol consumption.
The benefits were not discounted and the duration of benefits was not stated. Side effects of treatment were not relevant in this analysis.
Cost results At the current level of detection and intervention, the total cost was Aus$21,849,339.
In scenario 1, 5%, 10% and 100% increases in detection were associated with total costs of Aus$24,768,635, Aus$27,687,931 and Aus$58,385,924, respectively.
In scenario 2, 5%, 10% and 100% increases in intervention were associated with total costs of Aus$23,172,157, Aus$24,494,975 and Aus$43,638,128, respectively.
In scenario 3, 5%, 10% and 100% increases in effectiveness were all associated with total costs of Aus$21,849,339.
In scenario 4, a 5% increase in detection and intervention was associated with total costs of Aus$26,270,200. Similarly, a 5% increase in intervention and effectiveness and a 5% increase in detection and effectiveness were associated with total costs of Aus$23,172,157, and Aus$24,768,635, respectively.
The costs were not discounted.
Synthesis of costs and benefits The costs and benefits were summarised in the form of an average cost-effectiveness ratio by dividing the total costs by the total benefits for each scenario. Neither the costs nor the benefits were discounted. An incremental analysis was not performed.
At the current level of intervention and detection, the cost per patient modifying their drinking behaviour was Aus$231.45.
In scenario 1, 5%, 10% and 100% increases in detection were associated with cost-effectiveness ratios of Aus$232.60, Aus$233.51 and Aus$241.54, respectively.
In scenario 2, 5%, 10% and 100% increases in intervention were associated with cost-effectiveness ratios of Aus$191.26, Aus$165.61 and Aus$81.56, respectively.
In scenario 3, 5%, 10% and 100% increases in effectiveness were associated with cost-effectiveness ratios of Aus$208.31, Aus$189.37 and Aus$104.15, respectively.
In scenario 4, a 5% increase in detection and intervention was associated with a cost-effectiveness ratio of Aus$191.96. Similarly, a 5% increase in intervention and effectiveness and a 5% increase in detection and effectiveness were associated with cost-effectiveness ratios of Aus$172.14, and Aus$209.34, respectively.
Authors' conclusions From the perspective of the Australian Medical Benefits Scheme, increasing the rate at which general practitioners (GPs) offer an intervention to at-risk drinkers decreases the cost per patient reducing their alcohol consumption. To a lesser degree, increasing the effectiveness of the intervention also reduces the cost per patient. However, increasing the rate at which GPs detect at-risk drinkers is associated with an increase in the cost per patient.
CRD COMMENTARY - Selection of comparators It was not clear why the comparator scenarios used were chosen, and the authors did not provide a justification for their choice. You should decide if they represent valid comparators in your own setting.
Validity of estimate of measure of effectiveness Most of the evidence used in the decision model came from published studies. However, it appears that a systematic review of the literature was not carried out. Although this is common practice with models, it does not always ensure that the best data available are used in the model. Details of the primary studies were not provided and, in general, the values for the parameters were taken from a single source. Authors' assumptions were also used to derive one parameter, and the values used were justified on the basis that they represented a mid-point between values taken from the literature. These values were not tested in a sensitivity analysis.
Validity of estimate of measure of benefit The summary measure of benefit was the number of at-risk drinkers who reduced their alcohol consumption after detection and intervention by a GP. This was derived directly from the model. The decision analytic model used for this purpose was appropriate. The principal input parameters were derived from an ad hoc review and authors' assumptions, which might have biased the results of the health benefit measures obtained. Therefore, the quality of the input parameters is a limiting feature of the study. The measure of benefit was appropriate for the type of intervention under analysis but the authors did not discuss the duration of the benefits.
Validity of estimate of costs The cost analysis was performed from the perspective of the health service. It appears that all the categories of costs related to this perspective were included in the analysis. The costs were reported separately to other model parameters, thus enhancing the reproducibility of the study in other settings. The authors used a published source to obtain the costs of GP consultations. They assumed that detection and intervention by the GP would be reflected in longer consultation times. The costs were treated deterministically and no sensitivity analysis of the prices was conducted. Discounting was not applied, which was appropriate given that the model appeared to have a short time horizon. Costs, rather than charges, were reported, thus reflecting the true opportunity costs of the intervention. The year to which the prices referred was reported, and this increases the generalisability of the results.
Other issues The results of the analysis were not compared with the findings of other relevant studies. The issue of the generalisability of the results to other settings was also not addressed. The authors acknowledged four further limitations to their study. First, the rate at which at-risk drinkers consult a GP was assumed to be similar to that of the general population. Second, the definition of at-risk drinking varied between studies. Third, the rate of detection was taken from a published study and is likely to be an overestimation of real-life detection rates. Finally, the rates used did not reflect recent improvements in detection and intervention. The results were adequately reported and the authors' conclusions reflected the scope of the analysis. However, an incremental cost-effectiveness ratio would have been a better measure of the relative value of the scenarios instead of the average cost-effectiveness ratios.
Implications of the study The authors suggested that increasing the rate at which GPs provide interventions to at-risk drinkers achieves a greater benefit and return on resource use than increasing either the effectiveness of the intervention or the rate of detection of at-risk drinking.
Bibliographic details Doran C M, Shakeshaft A P, Fawcett J E. General practitioners' role in preventive medicine: scenario analysis using alcohol as a case study. Drug and Alcohol Review 2004; 23(4): 399-404 Indexing Status Subject indexing assigned by NLM MeSH Adolescent; Adult; Aged; Alcoholism /diagnosis /economics /prevention & Australia; Behavior Therapy /economics; Cost-Benefit Analysis /statistics & Family Practice /economics; Female; Health Knowledge, Attitudes, Practice; Humans; Male; Middle Aged; National Health Programs /economics; Patient Compliance /statistics & Physician's Role; Probability; Risk Assessment; Treatment Outcome; control /rehabilitation; numerical data; numerical data AccessionNumber 22005006159 Date bibliographic record published 31/03/2006 Date abstract record published 31/03/2006 |
|
|
|