[H+T--Friends] Health economic assessment tools (HEAT) for walking and cycling

Dear friends in H+T, There's a new health assessment tool, an online calculator for the economic benefits of increased walking and cycling, e.g., due to an intervention. Report: http://www.euro.who.int/__data/assets/pdf_file/0003/155631/E96097.pdf Calculator: http://www.heatwalkingcycling.org/ Having spent some time trying the calculator, I wrote the below overview and observations (including how to avoid certain pitfalls) to help people have an easier time with the tool. This forwarded message thread was originally written to the email discussion list of the Association of Pedestrian and Bicycle Professionals (APBP), in response to Anne Lusk at the Harvard School of Public Health. Jason ---------- Forwarded message ---------- From: Jason Meggs <jason(a)healthycity.net> Date: Wed, Jan 25, 2012 at 7:27 PM Subject: Re: [apbp] Health economic assessment tools (HEAT) for walking and cycling To: Anne Lusk <ANNELUSK(a)hsph.harvard.edu> Cc: members(a)lists.apbp.org Thanks Anne, Great stuff. We're experimenting with HEAT for the BICY project here in Central Europe. Very useful -- there's an online calculator: http://www.heatwalkingcycling.org/ This calculates, for either walking or cycling, "an economic assessment of the health benefits of walking or cycling by estimating the value of reduced mortality that results from specified amounts of walking or cycling". Even better, one can choose two points in time, when the amount of cycling changed, to show the benefits of an effort to increase cycling by calculating economic benefits over a period of years after the change.  I suspect one could also show the costs of some horrible change that made people walk or cycle less.  The tool can be more versatile than the instructions suggest. Because while experimenting with this calculator, I found it would be helpful to know what would be needed in advance (there are a number of paths to take, and a variety of data needed for each path), I am summarising it here for APBP folks. This is also because, *warning*,  it can be a bit buggy if one has more than one open at a time -- jumping to the results based on another of the previous/other instance(s) -- and possibly not functioning properly when one goes back and forth to try different options.  That's part of why I took the trouble of summarising the tool's input needs. I'll send this recommendation to the authors as well. Thus, if exploring with this, I recommend you open only one instance at a time and to start a new one, click on "exit the assessment" and/or close the one you have, and only then click on "start a new assessment". There are some limitations to be aware of in addition to the amount of literature available to base the calculator on, such as the target population being 20-64 years (no children, no seniors), and a focus only on regular trips (although you may be able to use irregular trips if you know the correct total and/or average amounts). The data required may also be challenging to obtain for many places and years. Although it gives great options by performing the calculation using either of three categories (trips, duration, or distance, and a fourth for walking -- number of steps). Thus it would be good if the project summarised this on the website before one starts, to help one choose the best path to work towards fulfilling (e.g., so you can plan based on whether the data is available). For the trips instance, one needs to provide the population of the regular cyclists as well as their average individual rates,  . For distance and duration, you need: * Amount per cyclist (minutes or hours for duration; km, metres or miles for distance) * Days cycled per year (always interesting to seek out weather patterns) * Number of individuals doing the amount of cycling entered For trips (average per person or total observed across a population), you need: average number of trips per person (adult), or total number of trips: A. average number of trips per person: - The average number of trips per person per day - Days per year do people cycle this amount - Proportion of these trips are cycling trips - Average trip duration or distance (WARNING, the default options are reversed here from the above; default is now hours not minutes for duration, and metres not km for distance) - The number of individuals doing the amount of cycling you entered B. Alternately, total number of trips: - the number of trips observed per day: - proportion of these trips are cycling trips - then either the number of people who take cycling trips, or estimate the number of cyclists based on the proportion of return journeys out of all trips observed (takes you to a calculator, with a page of instructions) - average trip duration or distance I did not fully assess yet whether walking benefits are calculated differently than cycling, but it looks the same with the addition of an exciting new option: C. Number of steps! For this you need: - Average # steps for an average day, week, month or year - Average step length After dealing with one of the three (or four, for walking) main choices above, the calculator is probably the same for the rest: You also need to choose a mortality rate, and the economic value of a life. (There are default values offered for Europe and some other countries, but not for the USA and Canada. You next choose the "standard value of a statistical life used in the country of study" (a problematic concept at best..) and the currency. Here you can use the CAD or USD, although CAD is not in alphabetical order. You can also assess the cost effectiveness of an intervention (you must know the cost of promoting cycling, and use a number for the expected change in cycling). So if you have an idea of how much cycling will increase from, say, a big promotional campaign, or building new facilities, you can make an economic assessment of the cost effectiveness of the intervention (although limited as you can only study part of the cycling, limited by age and whatever regular trips you have data for, but of course one can estimate or run various scenarios on possible outcomes). Lastly, you must also choose a discount rate (always an easy task!). A default of 5.0% is offered. Note that if you chose to compare two points in time, you will also enter your post-intervention cycling data (data on the second point in time).  Good news, you have three choices again in case your data types differ for the two points in time.  You will also be asked to estimate the proportion of the new cycling attributable to the intervention (or just 100% to assess all cycling, or to recklessly assume the intervention did all the work). You also have to give time to "maximum uptake" which basically is a fancy way of asking how many years between the two points in time.  (The calculator directions sometimes assumes too much; what if you are assessing two points that occurred in the past?)  Be sure to differentiate between that time period for the increase of cycling, and the time period during which benefits accrue (assuming cycling is now at a new steady level, total the benefits you get for X years.). I hope this is helpful and makes the calculator easier to use. I'd be interested to hear of anyone using these tools and what you think (and what your results were :). Happy calculating! Jason === Jason N. Meggs, J.M., M.C.P., M.P.H. University of Bologna DICAM - Transportation Engineering Group Viale Risorgimento, 2 40136 Bologna, Italy Office: +39 051 209 3338 Fax: +39 051 209 3337 Mobile: +39 333 1565 787 Email: jason.meggs(a)unibo.it           jason(a)healthycity.net           jmeggs(a)berkeley.edu Web: http://jasonmeggs.com/ Skype: jasonmeggs USA Phone: +1 510-725-9991 On Wed, Jan 25, 2012 at 5:06 PM, Anne Lusk <ANNELUSK(a)hsph.harvard.edu> wrote: