Avert (HIV and AIDS organisation)

Transcription

In segments 1 and 2, you learned why to do CEAs, and how to compare net costs to health outcomes, especially disability-adjusted life years (or, DALYs) and quality-adjusted life years (or, QALYs). In this segment, #3, we dive deep into DALYs, to understand how they're calculated and used. We also consider their close cousins the QALYs. We'll start with a nice video Let's queue up The DALY Show. Picture this: you're the Director of Chronic Disease Control in the Ministry of Health. You're meeting with your fellow directors, discussing health programs, when the Minister bursts in and announces that the Ministry has just been offered one hundred million dollars in additional funds. You think, so many of my programs could benefit from that type of money. However, there is a catch. The money would be given only to disease disease intervention strategies, which demonstratively offer high value for money. Or Bang for the Buck! That is the greatest health benefits for the most people with the available funds. And of course there are other agencies within the Ministry competing for this funding. The Minister proposes to focus on diabetes. A great cause, you think. As the Director of Chronic Disease Control, you know full well just how much diabetes affects peoples lives, but how do you make that argument in a way that others will understand? How do you demonstrate to the donor that diabetes compromises the health of many people, in important ways as much as diseases like, say, AIDS? After all, they're different in just about every way: physiology, complication rates, mortality risks, and how they influence overall health. Furthermore, how do you show that your diabetes intervention strategies and programs offer health benefits similar to or more than those of the competition? It's like comparing pineapples and bananas and broccoli. Wouldn't it be great if there were a single tool; something combining premature death and morbidity, that could comprehensively quantify disease burden; a single, universally recognized and respected unit of measurement you could use to evaluate all of these different disease factors?[3] Thankfully, there is! It's called a DALY, and it's exactly what you need! In technical jargon, it's a common metric that solves the pineapple-banana-broccoli, non-comparability problem. DALYs allow direct comparison and summation across diseases, as well as between untreated and treated disease. They also permit comparing different disease interventions, such as treatment expansion versus awareness and prevention campaigns, by comparing which intervention, to use the standard terminology, allows you to avert more DALYs. DALY stands for disability-adjusted life years. Simply put, a DALY is a standardized, quantitative measure of the burden of disease. A DALY score combines all the clinical effects of a disease with years of life lost due to premature death, or mortality into one valid, summary measurement. DALY estimates and calculations are divided into these same two parts: morbidity and mortality. Mortality is easy to quantify. It's simply the standard life expectancy of a healthy individual minus the age at death of an individual with the disease being studied. Think of this as the "LY" in DALY, or "life years" lost due to disease. Morbidity is a bit more complicated to calculate. First, each illness effect is given a severity rating, or disability weight, ranging from zero, which means fully healthy, to one, which means very severe. Next, since effects can be short or long term, the duration of the morbidity is factored in. Think of this as the "DA" in DALY, or "disability adjustment" due to morbidity.[8] When both mortality and morbidity contributions are added together, you have a full DALY score. Let's look at an example using adult-onset diabetes.[9] Let's say an individual dies at 60 years old, instead of the standard life expectancy of 75 years: that's 75 minus 60, so 15 DALYs.[10] Let's also say that the morbidity effects of adult-onset diabetes perhaps foot or kidney complications rendered this individual 50% disabled for the final ten years of his or her life. So, that's the severity rating, or disability weight, which we say in this case is 50%, so zero point five, times the duration, so 10, for ten years, results in 5 more DALYs.[11] So, for this individual, the overall disease burden due to diabetes can be represented as 20 DALYs: fifteen for mortality plus five for morbidity.[12] Usually we adjust DALY estimates for timing, meaning we treat future events as having less value from today's perspective because we aren't as concerned with the future as we are with the present. So, for instance, if we had adjusted the previous example according to timing, we would have calculated only 13 DALYs, not 20. This adjustment is known as discounting. Keep it in mind, but for now, we'll stick to exploring undiscounted DALYs. And what if a diabetes care intervention extended life by 5 years and reduced the duration of disability from 10 years to 4? Then the LY, the life years lost would be 10 instead of 15, and the DA, the disability adjustment, would be 2 instead of 5. Added up, that means we now have 12 DALYs instead of 20, which means, again, using the standard terminology, we've averted 8 DALYs through this intervention. DALYs were invented 30 years ago, to combine separate data on disease mortality and morbidity being recorded by the World Health Organization. It complements another common metric already in use at the time, known as QALYs or quality-adjusted life years. DALYs and QALYs both incorporate length of life and the negative effects of morbidity, but they're opposites. QALYs measure health so bigger values are better while DALYs measure disease burden and thus, are always undesirable, by definition smaller DALY values are better. Again, health programs try to avert DALYs. Other than it's opposite sign, DALYs and QALYs are almost equivalent. Since their introduction, DALYs have become the accepted way to quantify global disease burden, allowing observers to compare and evaluate disease data as grouped by country and region, age and gender, type of disease and more. For example, here you can see the distribution of DALYs by disease and age. In newborns, most DALYs are due to neonatal disorders. For one-month-olds, you can see diarrhea, lower respiratory illness, and other common infectious diseases cause the most burden. Diabetes takes a steady toll starting in the twenties. Cardiovascular disease takes off quickly from age thirty. Similarly, in this graph you can see how disease burden varies by geographic region. On the left are wealthier countries, with the largest burden from cardiovascular disease and cancer, and secondarily from musculoskeletal disorders, diabetes, and mental disorders. On the far right are African regions, with the highest burden from HIV, diarrhea and respiratory infection, tropical diseases, and neonatal disorders. You can find even more DALY-based graphics in a special, December 2012 issue of Lancet and online at the Institute for Health Metrics and Evaluation website. So, with DALYs, you can provide the Minister of Health with exactly the kind of estimates and information the donor wants to see. You can quantify the burden of diabetes and compare that to the burden of other diseases, such as AIDS. Granted, you would still need to collect data such as, how many people have diabetes, how much their lives are shortened on average, how much disability they suffer, and also gather the same data for individuals with HIV infection, in order to make these DALY comparisons. The necessary information is all generally available through clinical surveys, epidemiological studies and global burden of disease tables. You can even use DALYs to tackle the Minister's other challenge: quantifying and comparing different intervention programs addressing the same disease. You could calculate the number of DALYs a diabetes treatment program averted, perhaps by extending life or reducing the length or severity of morbidty, and then compare that with the number of DALYs averted by another program, perhaps focused on diabetes prevention through nutritional counseling. Then, by factoring in the potential one hundred million dollars in additional funds and the coverage which those funds allow for each intervention program, you can extend your analysis into a full-fledged, simple cost-effectiveness analysis. Certainly, this basic cost-effectiveness analysis is more detailed in practice but pretty simple conceptually in that you're chipping away at the DALY burden and comparing which option offers that attractive bang for the buck the donor wants.[20] DALYs are a big part of global health planning and evaluation. As you've seen, they offer a logical and flexible way to compare disease burden from different diseases, across populations, and with and without interventions. With the DALY concept and skill under your belt, you can understand and contribute to global health policy discussions. So, go forth and use DALYs in good health! So, now I'm going to say more about an issue only briefly mentioned in the video. That is -- how we quantify morbidity the disability weights used in DALYs and the utilities used in QALYs. The disability weight is how much people are disabled by their illness; the utility is how good they feel. Both are on a scale of 0 to 1, so a disability weight of 0 means no disability or fully healthy. Conversely, a utility of 1 means healthy. So both represent the relative severity of different diseases, disease stages, and disease events. You can compare how sick people are across diseases whether you're using disability weights or utilities. Having a limp and walking a little slower doesn't result in a big disability. Having chronic severe pain has a big disability. With disability weights or utilities, you can compare different diseases and problems. Both measures are also useful for to compare how well an intervention works in terms of improving health. So if someone has a severe pain and one intervention reduces half of that pain and another intervention gets rid of the limp completely, it may turn out that reducing the pain by half is more valuable than getting rid of the limp entirely because the limp didn't have such a bad effect on overall health. However there are some key differences between disability weights and utilities. Disability weights have the advantage of being derived from one process that was organized by the Institute for Health Metrics and Evaluation. Disability weights for several hundred diseases and health states have been estimated and are publicly available. When everyone is using the same numbers it tends to foster standardization and facilitates comparison across diseases, interventions, and studies. That's the plus side of disability weights. The plus side of utilities and QALYs is that there are far more measurements of it in scientific literature. Many people have measured utilities. The downside is that they've used different methods -- standard gamble, time trade off, visual analog scale, indirect measures. Also they've used different respondent groups, both with and without disease, which may have different perspectives. The measurement methods don't always agree with each other. Because the measurement values are different you may wind up with different utility values. This is best handled with sensitivity analyses. So you have a tricky choice when quantifying morbidity in cost effectiveness analysis. If you're going to use DALYs you use disability weights and if you're going to use QALYs you use utilities. The best place to go for standard disability weights is a paper published in the special issue of Lancet on the global burden of disease in late 2012. The first author for this article is Joshua Salomon. That pretty much wraps it up for DALYs and QALYs. Do you agree that they're uber? Feel free to watch this again. Most people need several iterations of these metrics to really understand them. In the next video segment, we'll see how CEA results can be presented.