Pharmaceutical Outsourcing Application
Pharmaceutical companies find it expedient to outsource some or all the work involved in running a clinical trial to CROs. This example looks at the distribution over different functions of costs in outsourcing budgets. In the other examples there was a hierarchy, levels in an organization, which spanned the vertical space of the graphic. Through this example we illustrate that such organizational hierarchies are not necessary, and, as noted elsewhere, any resource which can be represented in a two-way data matrix can be represented by the resource tepee. The costs of outsourcing are typically associated with the number of patients (N) that are to be enrolled in the study (though this association can be thrown off somewhat by special needs in the studies such as the collection and analysis of biomarker data, special study design features, clinical phase of the study etc.,). In this example the number of patients in the clinical studies being displayed is used to order the vertical space of the graphic. More generally one could create an ordering by using the row sums in the resource utilization matrix. The resource utilization matrix below shows the distribution of outsourcing costs over a collection of clinical studies.
The collection of studies could be defined as some meaningful collection such as one for those pertaining to a specific therapeutic area/drug or for a specific CRO. The idea is to develop a graphic which provides a holistic depiction of the distribution of costs in a slew of outsourced studies. The numbers in the resource utilization matrix that follows are made-up numbers (in $1000s) which bear no resemblance to any real outsourcing effort.
|Regulatory||Project Start Up||Project Management||Clinical Monitoring||Medical and Safety Mgt.||Data Management||Statistics/ Writing||Sub-contracting/ Other||Row Sum|
|Study 21, N=80||0||3124||2756||1700||190||479||417||204||8870|
|Study 07, N=160||0||7234||4124||2876||0||769||589||317||15909|
|Study 32, N=240||212||11031||6175||4100||287||879||669||354||23707|
|Study 12, N=280||314||12073||7245||4301||315||912||715||287||26162|
|Study 08, N=430||307||18756||11237||7568||324||1234||1079||371||40876|
|Study 03, N=510||289||21478||13421||9861||354||1267||996||425||48091|
|Study 04, N=580||0||23768||15123||11345||200||1450||1179||382||53447|
|Study 17, N=720||355||27456||17843||13289||423||1765||1289||415||62835|
The graphic below adds a null upper node and a null lower node. These nodes make the area of the colors in the graphic exactly proportional to the total cost, across studies, of the function associated with the color.