Conducting a clinical trial costs millions of dollars. Delivering a clinical trial portfolio may cost tens or hundreds, or even billions. When this amount of money is being invested in developing drugs then small percentages of improvement in productivity can have a major impact on profit and the ability to run more clinical trials for the same cost, and more significant improvements reap even higher rewards.

Key Highlights

Conducting a clinical trial costs millions of dollars. Delivering a clinical trial portfolio may cost tens or hundreds, or even billions. When this amount of money is being invested in developing drugs then small percentages of improvement in productivity can have a major impact on profit and the ability to run more clinical trials for the same cost, and more significant improvements reap even higher rewards.

Tweet
PRA Health Sciences
PRA Health Sciences

This article explains the principles and processes of accurate and timely workload and resource management – principles and processes that produced a 15% improvement in PRA’s clinical productivity. Productivity can be improved by many means but for our purpose we will focus solely on improving workload forecasting and resource management – for example, just by knowing that clinical monitoring work is required with enough notice to select a competent Clinical Research Associate (CRA) in the same city will greatly reduce travel time, travel cost and possibly accommodation costs too – imagine if the CRA could monitor for 2 days rather than 1 day in a particular week because of this knowledge – that prior knowledge is transformed into a 100% productivity increase.

Workload Forecasting

It takes 10-15 years to develop a drug and take it to market costing approximately $1.4B, and whilst a biopharma will have a drug development plan comprising of the clinical trials that will need to be performed to satisfy the Regulatory Agency’s questions it will not know exactly where and when those trials will take place. Whilst it is necessary to know what role will be performing which tasks where and when with a planning horizon of 6 months – to enable staff to be recruited and trained as necessary) – it is not necessary to know this for clinical trials that have not yet started. What is necessary is that the biopharma organization knows what budget will be required either for outsourcing the trial(s) or conducting the trials themselves, or utilizing an ‘embedded blended’ resourcing model where a combination of outsourcing and embedding is utilized to optimize the economics of clinical trial portfolio delivery. PRA has conducted thousands of clinical trials and for each phase and therapeutic area it has the real-life data to reveal the budget required per patient. This means that all the biopharma has to know is when the trials will start and stop, and how many patients will be involved, to be able to predict budget requirements. The exact budget required will never match original predictions, rather this process is used to broadly estimate what funds will be needed so that they can be sourced internally or externally.

The graphic below shows how the knowledge of the trials required, their start and stop dates, and the fees per patient can be used to create a total budget for the development of the drug – per year and cumulatively. In this example the drug development period is 12 years the total cost is $1.403,850B. The total cost is calculated by multiplying the fees budget per year by ‘3’ as drug development history shows that total cost which also includes ‘pass through costs’ (travel, accommodation, subsistence etc) and ‘investigator grants’ (contractual payments to sites for conducting the trial) is three times higher than budget fees alone.

Picture1 SSD Productivity Blog

Now that we know the budget requirements for the development of our drug we can focus on maximizing productivity through accurate resource forecasting.

How to Accurately Forecast Requirements

For many countries of the world due to the length of employee’s notice periods (can be 3 months), employment laws governing from when a resignation is valid, the recruitment cycle, and the on-boarding and training periods it is best to be able to predict the following with a 6-month planning horizon:

  1. The work to be performed eg: a monitoring visit and it’s accompanying tasks to include preparation, travel, trip-reporting and expense submission
  2. Which role will perform the work based on competence to do so
  3. Where the work will be performed – for larger countries the city or region of the country may need to be specified to reduce unnecessary travel commitments
  4. When the work will be performed – per calendar month is appropriate
  5. How many hours of effort the work will take
  6. How many hours are available for work in that month

The data table below shows how the information is presented so that the decision-making process for assigning staff to work is optimized. For example, in South Korea from July 2021 to November 2021 between 1.71 and 6.86 FTEs of CRA resource are required to perform the work.

Picture2 SSD Productivity Blog

This table is then aggregated for all clinical trials in the portfolio to reveal the number of CRAs required in each country per month. See the data table below – this is the demand for labor.

Picture3 SSD Productivity Blog

The demand for labor must then be compared to the supply of labor to understand where there is over-capacity and where there is under-capacity, and how over-capacity may be used to help the under-capacity. The data table below shows the FTE supply of CRAs employed by the biopharma in both countries for the remainder of 2021. Below is the data table showing the supply of labor.

Picture4 SSD Productivity Blog

The table below compares supply and demand by subtracting the demand number from the supply number. A negative number reveals a shortage of resource whilst a positive number indicates over-capacity. These are the variances which must be solved. Perfect resourcing occurs when there is zero variance and the entire workforce is fully utilized on value-adding work.

Picture5 SSD Productivity Blog

The final table (below) assists the resourcing process by coloring cells ‘green’ where the variances are <10%, and ‘red’ where they are >=10% variance. The reason for these ranges are that it may be reasonable in a country for the workforce to vary it’s intensity of work and/or it’s working hours for short periods to off-set the need to increase the supply of labor ie: employee goodwill supports the biopharma’s financial goals, and may lead to recognition of achievement by staff. So the good news is that France is balanced in Nov-Dec 2021 – the bad news is that for the next 4 months variances exceed 10% so there are decisions to be made. For South Korea the issues are even more serious as the under-capacity ranges from -96% (approximately half the labor required) to -380%.

Picture6 SSD Productivity Blog

Solutions to Resourcing Variations

There are four main categories of external solutions to under-capacity, and biopharma’s appetite for each will largely be driven by the economic benefits and desire to control the workforce rather than outsource work. The four types of solution are:

  1. Hire permanent staff. This is the best economic solution if there is a permanent shortage of staff. In the Korean example above it would be difficult to argue against the benefits of employing 1 additional FTE which would reduce the range of variances to +0.04% to -280%. The counter argument is that if workload reduces in South Korea in following years then the workforce has over-capacity and it may be challenging to re-deploy them or there could be legal and cultural resistances to ‘right-sizing’.
  2. Embed staff from a CRO vendor. This is the best economic solution if there is a permanent shortage of staff and the variability is relatively stable over time. In this way the biopharma organization optimizes cost by embedding staff in addition to it’s permanent staff who can perform all the work needed up the minimum level of FTEs required. This also provides biopharma with risk-free recruiting as they can benefit from transferring resource into their organization who are known and trained.
  3. Contract staff. This is the best economic model if the resource forecasting is volatile especially in shorter periods. The option exists for biopharma to hire and embed in predictable times and top up the capacity as needed. CROs such as PRA will offer this option too.
  4. Outsource work. This removes risks associated with hiring, embedding and contracting labor, but may not be the best economic decision as biopharma may be paying for infrastructure costs that they do not need, and there is the risk of contractual scope change leading to unpredicted Change Orders.

Productivity can also be greatly enhanced through internal solutions. Here is a step-wise thinking process which will drive up productivity. If the first or prior step does not solve the challenge then proceed to the next potential solution:

  1. If there is a variance, could the resource work a little harder or longer to complete the work with no additional resource assignment necessary?
  2. Is another competent resource available to perform the work that has spare capacity, and is preferably local?
  3. Could the work be performed remotely?
  4. Could a Manager and a more junior resource combine to perform the work?
  5. Could another resource from another function perform the work who has the correct competences eg: an ex-CRA who has recently moved into Clinical Trial Management?
  6. Could a Manager perform the work?
  7. Could a resource be trained in time to become competent?
  8. Could a competent resource from another country perform the work? For example, if a monitoring visit is required in Belgium and there is no Belgian CRAs available, could the work be performed by French or Dutch colleagues? Considerations include:

    1. Language
    2. Travel distance, cost and time – including length of time and cost of obtaining any necessary visas
    3. Length of assignment
    4. Training requirements
    5. Culture – including political histories between countries
  9. Could resource be ‘seconded’ from another country in another region of the world eg: French-speaking Canadian CRAs re-locating to France for a period, or Portuguese CRAs being re-located to Brazil?
  10. Try the external solutions listed above

Underpinning all these potential solutions are data and knowledge. At PRA our Anticipate and Shuffle workload and resource forecasting systems are programmed with the knowledge that is used in the internal solutions thinking process shared above. This has led to a 15% increase in productivity – not just by ensuring that all work has correctly assigned labor, but also that work that is no longer needed is removed from the system which releases labor back into the resource pool. PRA is happy to share the software on a commercial basis or to help biopharma develop it’s own versions.

Conclusion

Every industry desires higher productivity, and the buzzwords ‘innovation’ and ‘hacks’ are sprinkled throughout productivity conversation – and this is fine, but there is a but: successful business is also about doing the basics well, and successful resource management is the first place to start improving organization’s productivity.

You have challenges. We have solutions.

Start a conversation

Explore related content and topics