Differences

This shows you the differences between two versions of the page.

--- economic-evaluation-methods [2018/08/01 11:48]
thorsten [2.3.1 Economic acceptability criteria]
+++ economic-evaluation-methods [2018/08/01 11:54] (current)
thorsten [2.3.3 Quantifying surveillance benefits]
@@ Line 77: / Line 77: @@
 \\
 ===== 2.3.3 Quantifying surveillance benefits =====
-The important concept of loss avoidance is illustrated here with an example of early warning surveillance, where detecting disease early is expected to lead to a more rapid response (e.g. implementation of outbreak control measures) relative to the time of occurrence of the index case. Earlier in the outbreak, the losses already generated by the disease are smaller than at a later time. As a consequence, less spread means that the costs of intervention measures required are smaller than later in the outbreak with more animals and/or holdings being affected (Figure 3).\\ \\ {{/file/view/Fig3_.png/562258493/Fig3_.png|Fig3_.png}}\\ **Figure 3: Comparison of two surveillance options (S1 and S2) and their associated interventions (I1 and I2) in a situation where S2 leads to earlier detection of disease and I2 to effective disease control. The hatched area represents the losses avoided with the more effective combination of S2 and I2.**\\ To estimate the value of losses avoided, it is necessary to identify the effects in the animals or holdings affected as well as the effect of potential externalities. The losses can then be estimated by multiplying the number of animals of a certain type or species (e.g. dairy cows) suffering from a disease impact (e.g. reduction in milk yield) by the lost physical production coefficient (e.g. rate of reduced milk yield in dairy cows) and the price coefficient related to the disease impact (e.g. production price per litre cow milk). All disease effects need to be “translated” from a technical perspective into a value perspective in this way and summed up. The resulting difference in losses between two strategies is the benefit. \\ \\  For surveillance that aims to detect cases of a specified condition in order to implement an intervention, the same logic applies. Figure 4 illustrates two levels of surveillance with their associated interventions for three different baselines of decreasing, stable and increasing prevalence. In each case, the surveillance and intervention costs need to be estimated and compared to the estimated associated loss avoidance to determine the economic efficiency of the mitigation strategy. The strategy to be preferable from an economic point of view is the strategy that creates the largest net benefit, i.e. where the (positive) difference between loss avoidance and combined surveillance and intervention costs is largest.\\ {{/file/view/fig4.png/562258373/fig4.png|fig4.png}}**Figure 4: Comparison of two surveillance options and their associated interventions (S1&I1 and S2&I2) in a situation where the mitigation option 1 leads to disease reduction and the mitigation option 2 to disease elimination. A-c represent different baseline scenarios with decreasing, stable and increasing prevalence. The grey and hatched areas represent the losses avoided when the new strategies are compared to the baseline.**\\  Because the impact of surveillance cannot be measured directly as a mitigation outcome, it is only possible to quantify the loss avoidance resulting from the combination of surveillance and intervention and to compare it to the expenditures for surveillance and intervention. Therefore, it is recommended to calculate a residual margin over intervention cost which constitutes the maximum additional expenditures potentially available for surveillance without the net benefit from mitigation overall becoming zero. This margin can then be compared to the expenditures of various surveillance options and the one maximising the net benefit would be the best from an economic point of view. An illustration of this concept can be found in Häsler et al. (2012).\\ \\  Using intermediate outcome proxy measures for the benefit (e.g. number of cases detected) for inclusion in a cost-effectiveness analysis is only justifiable when a clear correlation between the intermediate outcome and the benefit is firmly established through empirical research. Otherwise, the case detection capacity of a surveillance system is non-interpretable from an economic point of view. A logic model can help to establish the relationships between intermediate and final outcomes. For example (Figure 5):\\ \\ {{/file/view/fig5.png/562258421/719x151/fig5.png|fig5.png}}Figure 5: A schematic logic model to illustrate the steps that leads from surveillance to a reduction of foodborne illness in humans.In the example depicted in Figure 5, if the aim of mitigation is to avoid human cases (the value of such avoidance can be quantified) and only the case detection capacity of surveillance is to be assessed and compared to the costs of the surveillance, the link between the intermediate outcome (disease detection) and the final outcome (reduction in human cases) needs to be established.\\
+The important concept of loss avoidance is illustrated here with an example of early warning surveillance, where detecting disease early is expected to lead to a more rapid response (e.g. implementation of outbreak control measures) relative to the time of occurrence of the index case. Earlier in the outbreak, the losses already generated by the disease are smaller than at a later time. As a consequence, less spread means that the costs of intervention measures required are smaller than later in the outbreak with more animals and/or holdings being affected (Figure 3).\\ \\
+{{ :files:fig3_.png?nolink |}}
+**Figure 3: Comparison of two surveillance options (S1 and S2) and their associated interventions (I1 and I2) in a situation where S2 leads to earlier detection of disease and I2 to effective disease control. The hatched area represents the losses avoided with the more effective combination of S2 and I2.**
+\\ \\ To estimate the value of losses avoided, it is necessary to identify the effects in the animals or holdings affected as well as the effect of potential externalities. The losses can then be estimated by multiplying the number of animals of a certain type or species (e.g. dairy cows) suffering from a disease impact (e.g. reduction in milk yield) by the lost physical production coefficient (e.g. rate of reduced milk yield in dairy cows) and the price coefficient related to the disease impact (e.g. production price per litre cow milk). All disease effects need to be “translated” from a technical perspective into a value perspective in this way and summed up. The resulting difference in losses between two strategies is the benefit. \\ \\  For surveillance that aims to detect cases of a specified condition in order to implement an intervention, the same logic applies. Figure 4 illustrates two levels of surveillance with their associated interventions for three different baselines of decreasing, stable and increasing prevalence. In each case, the surveillance and intervention costs need to be estimated and compared to the estimated associated loss avoidance to determine the economic efficiency of the mitigation strategy. The strategy to be preferable from an economic point of view is the strategy that creates the largest net benefit, i.e. where the (positive) difference between loss avoidance and combined surveillance and intervention costs is largest.\\ \\
+{{ :files:fig4.png?nolink |}}
+**Figure 4: Comparison of two surveillance options and their associated interventions (S1&I1 and S2&I2) in a situation where the mitigation option 1 leads to disease reduction and the mitigation option 2 to disease elimination. A-c represent different baseline scenarios with decreasing, stable and increasing prevalence. The grey and hatched areas represent the losses avoided when the new strategies are compared to the baseline.**
+\\ \\ Because the impact of surveillance cannot be measured directly as a mitigation outcome, it is only possible to quantify the loss avoidance resulting from the combination of surveillance and intervention and to compare it to the expenditures for surveillance and intervention. Therefore, it is recommended to calculate a residual margin over intervention cost which constitutes the maximum additional expenditures potentially available for surveillance without the net benefit from mitigation overall becoming zero. This margin can then be compared to the expenditures of various surveillance options and the one maximising the net benefit would be the best from an economic point of view. An illustration of this concept can be found in Häsler et al. (2012).\\ \\  Using intermediate outcome proxy measures for the benefit (e.g. number of cases detected) for inclusion in a cost-effectiveness analysis is only justifiable when a clear correlation between the intermediate outcome and the benefit is firmly established through empirical research. Otherwise, the case detection capacity of a surveillance system is non-interpretable from an economic point of view. A logic model can help to establish the relationships between intermediate and final outcomes. For example (Figure 5):\\ \\
+{{ :files:fig5.png?nolink |}}
+**Figure 5: A schematic logic model to illustrate the steps that leads from surveillance to a reduction of foodborne illness in humans.**
+\\ \\ In the example depicted in Figure 5, if the aim of mitigation is to avoid human cases (the value of such avoidance can be quantified) and only the case detection capacity of surveillance is to be assessed and compared to the costs of the surveillance, the link between the intermediate outcome (disease detection) and the final outcome (reduction in human cases) needs to be established.\\