Up-down sequential allocation

Yesterday, a fellow graduate student and a reader of this blog 'S' requested information about a statistical tool called 'up-down sequential allocation'. I thought, I will answer this as a post in my blog. I am doing this for two reasons. One, it may help others who are looking for the same information and also at the same time, I can maintain the frequency of my blog posts.

'Up-down sequential allocation' is a method which was initially used in social sciences and later found a great deal of usage in Anaesthasia studies, Inhalational agents studies. In fact a variation of it has been a favorite of brokers in picking up shares in stock-market.

It is now clear that most previous research has been above the 95th centile for efficacy for concentrations (EC95) and doses (ED95), at the top of the concentration or dose-response curve where subtle differences will be missed or apparently similar efficacies misinterpreted. (Ref: Dr. Malachy Columb, in Regional analgesia in Obstetrics; A Millenium update)

There has been a trend in the recent past to reduce the dose of epidural anaesthetic agents either by using alone or adding them with other anaesthetic agents like Opioids which may have a synergistic action. If you look for a pubmed search, you will get surprised by the sheer number of studies done using this approach.

In order to evaluate concentration response curves for anesthetic agents, the minimum local analgesic Concentration (MLAC) model was designed. This would allow the determination of the Median Effective Concentration (EC50) of local anesthetics in a clinical scenario.

Let us check this with one example. If you r looking at the difference between drug 'A' and drug 'B' in epidural analgesia and their combination. Initially allocate the subjects to three different groups. Now for the dosing, within the group, use 'up-down allocation'. First patient in that group will get an arbitrary dose of the agent. say some 10 mL. That patient may respond with 2 different responses. Effective or Not-effective. The response will be quantitated using a VAS(Visual Analog Scale) score. Also look for the rescue and the quantity of the agent required. Then the dose to the next patient depends on the response of the earlier patient. The dose will be calculated by using a statistical formula for up-down sequential allocation given by Mossey. This ensures all your doses will be with in the median range of effectiveness and with in the specified confidence intervals.

That is the basic funda. Now many researchers are using some variations in this concept to increase the usability. I know, it is bit difficult to understand the concept. Once with this basic knowledge, if you go through a research paper, I am sure you can get the picture with clarity.

Let me know if you have any queries on this. When I get free I can explain the stuff with a proper live example. Leave your comments.

Seeji, Pharm House


  1. Anonymous says

    It is really a very useful article. The author described very clearly the method of up-down sequential allocation which offen used in clincal study such as local anethetics potency.
    My big thanks to the author for his introduction of such a useful method to us.

    Seeji says

    Thank you Chen,

    I am glad that it was useful. Keep visiting.