IRT Configuration (2)

Table of contents

• Shipment triggering rule
• Shipment validity rule at site
• Shipment validity rule at depot
• Sites/depots resupply strategy
• Minimum buffer update period for sites
• Minimum buffer update period for depots
• Missed visit cost
• Out of stock cost
• Visits for stability

Shipment triggering rule

This parameter determines how the stocks are compared to the buffer level for shipment triggering. Two options are available :

• At value means that a shipment will be raised when the stock reaches the trigger level (min bufffer level) or gets lower.

• Below value means that a shipment will be raised only when the stock gets strictly below the trigger level (min buffer level).

Shipment validity rule at site

You specifies, at a site level, what should be optimized in priority: the overage or the number of shipments or a combination of the two.

Before going into details, we must define what constitutes a kit that is valid for shipping. Basically, a kit is valid for shipping, based on its expiry, if it can still be used once reaching the site. Assessing if a kit is valid for shipping will take into account the kit life (how long the kit should still be valid after dispensing), the site shipping lead time and the minimum storage period (MSP). The minimum storage period (MSP) is defined in the Resupply management tab of the IRT setup.

The Min wastage option will always favor the oldest expiries to be sent at site as long as the expiry is sufficient for the kits to be considered valid for shipping. This gives the best chance for all kits to be used, hence decreasing the waste.

The Min shipping option will aim to anticipate a bit the expiry replacements. Starting from one full long window before kits become invalid for shipping, the algorithm will try to already send kits with a longer validity. If this constraint cannot be matched (if there is only one lot available at the time, for example), other kits can still be sent as long as they are still valid for shipping. This option gives the best chance of having smooth expiry replacements and reducing the number of shipments around them.

The Mean ratio option does not relate to how actual IRT algorithms work so it should not be used as a strategy for site management.

Shipment validity rule at depot

You specify, at a depot level, what should be optimized in priority: the overage or the number of shipments (or a combination of the two).

Before going into details, we must define what constitutes a kit that is valid for shipping to depots. Basically, a kit is valid for shipping to depot, based on its expiry, if it can still be shipping to sites when reaching the depot. Assessing if a kit is valid for shipping to depot will take into account the do not ship (DNS) date (how long before expiry the kit is unable to be sent at sites, see above), the depot shipping lead time and the minimum storage period (MSP) at depot. The minimum storage period (MSP) is defined in the Resupply management tab of the IRT setup.

The Min wastage and Min shipping options will work exactly as they do for sites (see above). As a reminder:

• The Min wastage option gives the best chance for all kits to be used, hence decreasing the waste.

• The Min shipping option gives the best chance of having smooth expiry replacements and reducing the number of shipments around them.

The Mean ratio option works quite differently. The idea behind this algorithm is to optimize the ratio between kits about to expire and newer kits at depot, at all times. For every day that is simulated, a ratio will be assigned per depot and package type, based on the different lots available that day. For example, depot US on May 13th, 2019 could have a ratio of 80% kits with expiry in March 2020 and 20% kits with expiry in July 2020.
The ratios will be computed based on what is observed in the simulations. Every time a kit is sent from a depot to a site, the algorithm will save in memory what was the expiry of that kit. Aggregating this over the whole set of simulations, it is possible to know every day how likely a depot is to ship out kits with the different expiries available at that time. This likelihood is used to compute the ratios at depot, which will be different per depot and package type.

Sites/depots resupply strategy

You specify if the resupply of kit types should be done for all kit types of the same temperature types at the same time or individually.

The All products option means that when a kit type reaches its trigger level (min buffer level), all kit types from the same temperature types will be resupplied to the their max level (resupply level).

The Products at minimum option means that when a kit type reaches its trigger level (min buffer level), only this kit type will be resupplied to its max level (resupply level).

Minimum buffer update period for sites

If you do not define buffer levels (min/max or trigger/resupply levels) at sites, they will be computed during the simulations by the CT-FAST engine. The algorithm is able to compute the most suited buffer levels every day, however this is not a viable option as it would mean that these values have to be updated every day in the IRT system.
This is why you can constraint the engine to only update the buffer levels after a certain amount of time has passed. If the refresh period is set to 90 days, the algorithm will have the opportunity to change the buffer levels only every 90 days.

Minimum buffer update period for depots

As explained just above for sites, you can constraint the update of the resupply settings at depot. By default, the CT-FAST algorithm will update the trigger and resupply levels at depot every day.

Missed visit cost

A cost can be assigned to any event where a patient comes to the site but cannot receive all the kits that have been allocated to him/her. It is a way to represent that having a patient missing a visit can have costly consequences (for example, the need to re-schedule a visit).

Out of stock cost

A cost can be assigned to supply chain stock out events, as they are critical events that need to be dealt with by both the supply and clinical teams.

Visits for stability

This last option will determine how the patient dispensing prediction will work. To simplify a bit, the prediction is quite simple: the CT-FAST algorithm will predict the needs for the patient's next visit(s) based on the dose level the patient is on. This allows to always have on site the kits necessary for patients that stay on the same dose level (titrations will typically be covered by buffers).
However, a question has to be addressed about what to do when a patient just titrated. In most cases, the prediction will adapt right away and predict for this patient based on the new dose level. But if there is a high chance that this patient might titrate again directly, you might want to stop predicting until the patient shows stable behavior again.

The number of visits for stability is exactly defined to tackle this question. You will enter how many times patients have to be stable (remain on the same dose level) before the prediction is applied to them.