Faster PCR results without adding staff or equipment

Summary

• Reorganising existing staff during busy periods reduced average PCR turnaround from more than 48 hours to about 32 hours in one simulated scenario, without additional hiring or equipment.

  
  

• How laboratories organised their work mattered as much as capacity: an inefficient staffing schedule pushed waiting times beyond 200 hours, while a five-day shortage of essential testing materials created a backlog that lasted for weeks.

  
  

• Targeted technology could reduce delays further. During high-demand periods, introducing barcoding brought reporting times down from around 70 hours to under 48 hours.

  
  

Why can a laboratory have capacity and still produce long delays?

A PCR sample passes through several linked stages, including registration, RNA extraction, amplification and result reporting. A bottleneck at one stage can delay the entire process even when other parts of the laboratory have spare capacity.

In Nepal, many laboratories outside the capital relied heavily on manual processes and had to manage changing sample volumes, staffing constraints and periodic shortages of essential testing materials. Evidence on how these conditions affect turnaround was limited because much of previous operational research had mainly focused on highly automated laboratories.

Why is this a prescriptive analytics problem?

The operational question is what should change when demand rises: how staff should be assigned, what schedules should be used, how much reagent stock is needed and where automation would have the greatest effect.

These choices interact. Adding equipment may have little value if another stage remains the bottleneck, while moving existing staff at the right time may improve flow without additional investment. Simulation allows these decisions to be tested before changes are made in a live laboratory.

How did the team model laboratory operations?

The study brought together CIRRELT and Polytechnique Montreal, George Washington University, Karlsruhe Institute of Technology, Amsterdam Business School, Nepal's National Public Health Laboratory and Ministry of Health and Population, the World Health Organisation country office and South-East Asia Regional Office, and the UK Health Security Agency.

The team built a discrete-event simulation that represented each stage from sample arrival to result reporting. The model used expert interviews, daily Ministry of Health situation reports, and an equipment and staffing survey covering 77 laboratories.

They then tested changes in staffing and schedules, shortages of essential testing materials with different durations and automation options such as barcode scanners and robotic sample handling.

What did the comparison show?

In one scenario, reallocating existing scientific staff to registration and data entry during busy periods reduced average turnaround from more than 48 hours to about 32 hours without hiring additional staff or purchasing new equipment.

Scheduling also created substantial downside risk. A poorly designed schedule pushed waiting times above 200 hours. A one-day shortage of essential testing materials had little effect, whereas a five-day shortage created a backlog that required weeks to clear. In high-demand periods, barcoding reduced reporting time from around 70 hours to under 48 hours.

For people waiting for a diagnosis, shorter and more predictable turnaround can reduce the time spent uncertain about whether to isolate, seek treatment or return to work or school. The findings also show that operational changes can be evaluated alongside new investment, rather than assuming that additional infrastructure is always the first response.

How can the findings impact other applications?

This modelling approach can be applied to other diagnostic laboratories that need to decide how to use constrained staffing, equipment and supplies under variable demand. It can also inform preparedness by showing how long a laboratory can absorb a supply interruption before backlog becomes difficult to recover.

Future applications could compare targeted automation with changes in staffing and scheduling, allowing laboratories and public-health authorities to identify which intervention addresses the actual bottleneck in a specific setting.