The first instalment of this blog looked at methodologies and documentation related to maintenance and calibration of laboratory equipment. In this second part we look at management of the processes of methodology and documentation.
The ultimate goal for any IVF Unit is a safe laboratory that achieves high pregnancy rates. The challenge is keeping that goal within the boundaries of profitability and sustainability. Data acquisition and data management plays a huge part in this process. This plethora of data should:
– Be easily accessible
– Highlight problems with visual presentation such as graphs with red lines
– Be regularly updated
– Involve more than one person at critical performance measurement junctures
By introducing Laboratory Performance Measures (LPMs) this data is a useful management tool for risk mitigation in IVF laboratories.
What is a Laboratory Performance Measure?
An LPM is an expression of data that shows what is actually going on in the laboratory. Quite simply, it is a data set that is determined in order to show an expected development stage or output. The process of performing LPMs can be considered to be analogous to continual sampling in a production manufacturing business. LPMs follow data from specific groups of patients to ensure that the laboratory is performing within expected parameters.
LPMs can be set up for:
- Quality markers: oocyte maturity, oocyte morphology, average egg numbers etc
- Culture quality: blastulation rate, survival rate of thawed embryos, utilisation rate etc
- Technical endpoints: fertilisation rate
- Toxicity markers: 2PN cleavage ratios
- Successful outcomes markers: pregnancy rates
LPMs allow you to choose 5 to 7 points of data that covers the main activities of the laboratory. If you have access to the data from more than one laboratory you can compare the data and benchmark performance.
The traffic light system
Once LPMs are chosen, you can what constitutes Achievement, Action and Warning.
When teaching, I would discuss half-year tests with pupils to determine their likely performance in the end-of-year exams. Some would be on course to achieve, some needed to take action and some had a warning of failure!
A visual traffic light system within the reporting system designates the level of performance: red, amber and green lights.
GREEN LIGHT
Green is for go. It means that all the LPM data is acceptable. But how do we know this if we are just starting out? You can use existing data, historical data or even AVERAGE data to set an achievement benchmark. The trick here is not to make the benchmark too high to achieve or to make it aspirational. It should be the middle rung of the ladder of success. The aim is to improve upon it by examination and adjustment of procedures that improve the LPMs. This means that improvements are evidence-based and data-driven.
AMBER LIGHT
This is a situation where one or more critical LPMs are performing too close to the red-light level. At what point should you take action? This will depend on the type of LPM and also on the data sampling frequency. For instance, you may want to wait until the next data comes in before making a decision. If the LPM is critical, action should be taken immediately to influence the next set of data positively.
RED LIGHT
A red light is a warning light. Simply, it is an LPM value that you don’t want to drop below. It is a call to action to analyse the problem and remedy it quickly. You can liken it to a car engine light, it necessitates pulling over when safe to see why it has lit up.
Pregnancy data
This issue of data sampling frequency pertains particularly to pregnancy data. LPM markers referring to egg and embryo development can be obtained weekly. But pregnancy data refers to patients and so several weeks data often needs to be combined. This ensures that the cycle numbers are high enough for the data to be meaningful.
Data reviews
The latency of pregnancy data means LPM markers for toxicity and quality will drop before the pregnancy LPM markers. Therefore, corrective action can usually be taken before problems affect the productivity of the entire business.
This, of course, is the aim of putting risk reduction strategies in place and using self-generated data to monitor them. During reviews of this data it is important to take a broad view and consider:
i. Data sample size: is the sample large enough to iron out idiosyncrasies that can occur with small samples?
ii. Benchmarks: can you compare your data with another laboratory?
iii. Cycle type: is the data segregated by first cycle versus consequent cycles?
iv. Pregnancy data: are you looking at like-for-like in terms of process? Are transfers all the same type? And are insemination methods similar?
v. Staff: which staff are allocated to which procedures that affect LPMs in the data period? Were trainees assigned to any key tasks relating to LPMs in the period?
Taking into account the above considerations, red lights need immediate investigation. Amber lights can usually wait for the next set of data before action is taken.
Laboratory review
If there are too many red lights, then a laboratory review may be indicated. This would look at the operation, the equipment, and all the technical processes in the laboratory. Examples include (not exclusively):
- Temperatures of all necessary stages in the egg collection area
- Incubators
- Egg collection procedures (e.g. suction pressures)
- Media/oil batch numbers
- Checking of gas cylinders
- Checking culture lab ware
It is important to note that LPM may still be negative if the data sample is too small to mask idiosyncrasies.
Running a laboratory that has instituted data-driven risk mitigation strategies, allows you to act proactively rather than reactively. It means there should be no surprises. It is one of the key reasons why well-managed, low-risk laboratories rarely experience significant and unknown pregnancy rate reduction.
Neil Madden, Editor
The Fertility Hub
