MellowoodMedical Online Help
LPI query - validation of source data and investigation of results - Worked Examples
- Mhairi West
Query outputs will reflect the data entered into lab events in the treatment cycles. Validation of this data is important as the report will only be as accurate as the data that is entered. Unusual results seen in LPI graphs should always be investigated to determine if they are a true reflection of lab data, or if there have been some data inaccuracies or omissions to skew the results.
In order to this you can always browse the source data, that supplies the summary graphs seen in the LPI report.
Browse LPI Query data
To see the report data in a table click on Browse.
Fill in the data range.
This will open the browse form, which contains all the data which has supplied the summary graphs seen in the report.
Exporting the Browse data
To export the contents click on the Export button. There are two options;
File | This will save it directly into an Excel file in the export folder. When you open it, it will look like this: |
 | |
Microsoft Excel | This will open Microsoft Excel. The columns are lined out, so this is the recommended option. |
 | |
Example - Investigating report results/validating source data - fertilisation rates
E.g in the IVF fertilisation rate graph, the fertilisation rate cycles appears to have dropped significantly for one month
The raw data for this month can be seen in the table below - the fertilisation rate had fallen to 38%.
The user may want to investigate the cause of this drop. It could be;
incomplete/incorrect data entered into treatment cycle lab events
a true fall in fertilisation rate
If it is the second option, the user may want to identify those exact treatment cycles and look e.g at the media in use that month, the average age of the patients, the quality of the insemination sperm, an operator related issue.
In the IVF graph you can see there is a result that falls out of the average range.
In the table you can see the number of IVF inseminated eggs this month and the % of success.
To validate/investigate the source data click on Browse, click on OK and export to Excel.
You can then add filters to the column headers and use them to select
cycles started in the month identified in the graph
Further filter to IVF cycles
This will produce the group of patients (showing their eggs and their fates line by line) that should be investigated further.
Initially you should look for potentially missing or incorrect data e.g the fertcheck lab event may accidentally have been entered on day 2 - thus indicating no fertilisation on day one.
Once the data validation is done, run the report again to see if the low fertilisation result is still indicated. If so you will need to investigate that group of patient further,
For help on the column headers in the Browse data sheet, and how the summary stats are calculated, click here LPI BROWSE data - column headers
Example - Investigating LPI query results - practitioner data
E.g how to look at results per IVF operator.
Open the Sort and group dialog.
Select ‘ICSI/IVF operator’
In this chart you can see that 2 IVF operators are below average.
There was a low average fertilisation rate in eggs inseminated by Isabel (orange line) in July and November, and by Siobhan in November
You also can isolate the individual IVF operator, by selecting 'compare average with' and choosing that operator from the drop down.
This report graph shows that she performed below the boundaries of 2 and 3 standard deviations.
In the table you can see the results in numbers, from Isobel’s cases compared to the group as a whole
Again, as in the example above, you will need to investigate whether there is data validation issue or this is a true result. This can be done by browsing and then filtering the data. To identify the relevant cases;
filter to show Cases with November and July start dates
Filter to show IVF
Filter to show cases where inseminating embryologist is Isobel
this will produce the source data behind the summary graphs and tables in the report