Typical Data Requirements

We have worked very hard to limit our data requirements to those data sets absolutely essential to supply chain modeling and improvement. We do not collect data just to collect data. If we don't use it, we don't ask for it.

Data Privacy and Security

Data privacy and security are top concerns of nearly everyone in management these days. So we will deal with this critical issue at the outset:

• Your data is secure. We store data in encrypted format except when modeling or analysis is being done.

• Your data is kept absolutely private. We do not share your data with anyone. Period.

• Your data is deleted as soon as our work is done unless you decide on continuing services. We use a multi-pass, overwriting deletion utility that really, truly deletes data.

Data List

Here is what we typically ask for in the initial model setup and action plan process:

1. Item master file
2. Locations master file
3. Unit of measure table
4. Purchase order lines table
5. Central supply inventory snapshot
6. Point-of-use storage inventory snapshot (if available)
7. Receiving lines table
8. Central supply issues and returns transactions file
9. Hospital staffing by unit (position, FTEs)
10. PAR and refill settings by stock location and product
11. Point-of-use transactions (if available)¹

We try to get 3 months of transactions (purchase order lines, receiving lines, stores issues and returns, and point-of-use transactions if available). This sufficient to capture the most important order patterns and to provide a basis for demand trend and cyclicality modeling.

On-Site Work

We normally also interview a number of hospital staff who handle supplies to get insights into issues and real-world mechanics. From these, we learn such vital insights as ...

• Backorders are a problem not because of their frequency (low) but because of the ineffective handling of backorders by our primary vendor.

• Unusually large supplies "takes" are often caused by nurses from distant locations on their floor trying to reduce the number of time-consuming trips to their unit's stockroom.

Data Transmittal

In most cases, data volumes are much larger than most email systems permit. We normally ask for data to be transmitted by CD, DVD or USB flash drive via a documented package service such as FedEx or USPS Express Mail. We prefer not to use FTP transfers because of the difficulty securing the data in transit and ensuring that the transfer goes only where it was intended. Data should be password-secured as a minimum.

Practice:   2. Cost Measurement

In addition to transaction and inventory data, we must have a clear picture of costs involved in supply chain activities. This is not a minor task the first time though but the effort is vital ... Measuring Supply Chain Costs ...

To return to the main topic list ...

 

¹ Point-of-use (POU) transactions are generally available only where automated supply stations are installed. Data extracted from these systems can be extremely valuable in demand characterization and forecasting and ordering process management.

 

Data Preparation and Use

Your data set is used for two main purposes:

1.  Model setup and validation

2.  Improvement simulations

We load your data into a convenient database application —usually MS SQL Server— and carry out a number of data quality and integrity checks.

These checks help identify item IDs that are used for special orders such as blanket orders and rentals or that may refer to more than one item description. We also flag items without description, unit of measure, vendor and similar information.

Because we treat the data set as a very large but possibly incomplete sample, dropping a small percentage of items and transactions is not normally a problem.