October 2002
When is a Page Not a Page?
A personal impression of Data Management in Japan
Having spent a year in Data Management in Japan, this is a good moment to sit back and reflect on the differences and similarities of data management between Japan and the rest of the world and the challenges of working in this environment.
The first very obvious difference lies in the answer to the question above – when it's a Japanese page. The amount of data that is on a Japanese page is about three times that which you would have on European page. Can you imagine having concomitant medications, adverse events and laboratory data ALL on the same page? One's worst nightmare come true! Added to this many CRFs have much more text than in Europe. This gives us problems in data entry as the kanji can sometimes be difficult to read (doctor's handwriting is the same wherever you are!). The text is not double entered but it does have to be checked visually to make sure the characters are correct. The fact that the pages are different gives me a problem in using our worldwide pricing model as the metrics and algorithms are clearly linked to an European/US style page. So I have to remember to multiply by 3. It gets very complicated when we try and compare our metrics across our other offices because my figures look completely different. But one gets used to the arithmetic.
Another major difference is the concept of 'booklet' versus 'visit' type CRFs. A visit-type CRF is one designed as Europeans know it: i.e. the pages can be pulled visit by visit and processed in an on-going fashion. I do remember though, in my previous company, that we discovered that 'visit by visit' collection meant the CRA's visit not the patient's! For this type of CRF, we use Data Clarification Forms to do the querying and we do QC in the usual way. A booklet type is the traditional CRF where the booklet stays with the investigator whilst a photocopy comes in to be processed. The queries go out and the original CRF is corrected. Eventually when it is declared clean, the original comes in and Third Entry takes place. This is unique to Japan, but in fact it is a QC step as the idea is that the complete, corrected CRF should match the clean database. The disadvantage of this system is that the data tends to come in towards the end of the study, so that locking a database in 30 days from last CRF in is just impossible. Six months is not uncommon!
Then there is the issue of kanji systems. Clintrial has been available in kanji for a few years now, but Oracle Clinical is only just now becoming available. This means that many companies still use homegrown systems. Some of those whose parent company insist they use OC have to translate everything first, which is not very efficient if the data is only needed for the Japanese market. In addition to this, Japanese companies have not been used to all the regulations such as Computer System Validation and 21 CRF Part 11, so there is a lot of catching up to be done, particularly for in-house systems.
I suppose you all think, as I did, that with ICH we now have one standard of GCP! Wrong! Japan still has its own version of GCP, but I have to admit that even though I have asked many times, I have yet to discover from a data management point of view just what the difference is. Once you remember that in the ICH GCP data management is hardly mentioned it's difficult to see where the difference really is. The other thing that is the same, yet different, is MedDRA(J). However this is due to the difference in medical concepts and terminology. Currently we feel the best way of translating adverse events into English is to do it via the decode, rather than risking mistranslating the verbatim term.
Data management in Japan is still a young profession, which means it is extremely difficult to find experienced staff. Traditionally the Japanese have not moved from company to company, though that is becoming more common now with younger people. In Quintiles we have grown extremely fast over the last few years so not only have we been short of technical expertise, but also people with sufficient people management skills to lead teams and deal with clients. However it does mean we have a young and very enthusiastic staff, who are also extremely dedicated to their job and their clients.
Managing staff in a different culture and language is always a challenge. At least when I worked in France I could speak the language! Here, although I am learning Japanese, I have to work through my PA who translates for me. I have learnt to speak in short bursts and to keep my train of thought whilst it's translated. However I probably do miss nuances and we sometimes do get completely the wrong idea. The letters 'r' and 'l' get very mixed up. There is somewhat of a difference between collecting a CRF and correcting a CRF! That one took a little while to untangle. The cultural aspects are more difficult to pick up but they are very patient with the gaijins. Business cards are a must and anyone coming to Japan for a business trip is well advised to get their card translated. In an emergency most of the big hotels will do it for you.
All in all it's been a very interesting year and I have learnt a lot. If anyone should get the chance of a year's secondment to Japan, they should seize the opportunity. Apart from the challenges of the work, it is a beautiful country and of course culturally very rich. We get the odd typhoon, the World Cup and little earthquakes from time to time but it is a very pleasant place to live.
Alex Paton
Executive Director, Data Management,
Quintiles Japan