Received: 19 April 2017
Revised: 18 July 2017
Accepted: 18 July 2017
DOI: 10.1002/pds.4290
LETTER TO THE EDITOR
Post‐marketing surveillance in Japan: Potential best way forward
Another problem is that the majority of “drug use‐results surveys”
To the Editor, Post‐marketing surveillance (PMS) in Japan aims to evaluate the
are not published, although the number has increased in recent
safety and effectiveness of newly marketed drugs based on a ministe-
years. Indeed, recent research has revealed that physicians are not
rial ordinance, the Good Post‐marketing Study Practice (GPSP), and its
satisfied with feedback from the results of PMS studies.4
related laws and regulations. Under the GPSP, pharmaceutical compa-
The cost of conducting PMS is enormous. According to the
nies must make provisions to conduct the traditional “drug use‐results
Transparency Guideline of the Japan Pharmaceutical Manufacturers
survey.” Data from this survey form part of the re‐examination
Association,5 pharmaceutical companies in Japan disclosed annual
application documents that must be submitted 6 to 10 years after a
remuneration to physicians for “drug use‐results surveys” of
medicinal product is first approved. The Japanese Society for
approximately JPY 13.1 billion in 2015. For example, at Takeda
Pharmacoepidemiology (JSPE) established the “Task Force for Good
Pharmaceutical Co. Ltd. (Takeda), the annual remuneration was
Pharmacovigilance Planning in Japan” in 2015 and highlighted the pit-
JPY 0.31 billion in 2015 (disclosed according to the Transparency
falls of the use of this “drug use‐results survey” as current PMS prac-
Guideline). The total costs at Takeda in 2015 were calculated to
tice in Japan. Problems include its indiscriminate use and lack of any
be JPY 3.05 billion, which included direct and indirect costs so
requirement for a comparison group or clear research question.1 The
far not publicly disclosed (see Supporting Information for the details
JSPE also suggested that the current GPSP should be completely
of the calculation). Therefore, the ratio of total cost to remunera-
revised to encourage reference to the annex “Pharmacovigilance
tion at Takeda was estimated at 9.8 (3.05/0.31). We estimated
Methods” in the ICH E2E guideline “Pharmacovigilance Planning”
the total cost of “drug use‐results surveys” in Japan as the remu-
(issued as a notification by the Ministry of Health, Labour, and Welfare
neration to physicians in the nation (JPY 13.1 billion, publicly
[MHLW] in 2005), which indicates that there are various ways to best
disclosed) × the ratio of the total cost to remuneration. According
address a research question.1
to the sensitivity analysis where the ratio was varied around the
PMS that does not meet scientific standards are not unique to
value at Takeda (9.8), the total cost in Japan in 2015 was JPY
Japan. For example, Spelsberg et al's survey of 558 notification
100 to 150 billion (approximately USD 0.88–1.32 billion) (see
documents submitted to regulatory agencies suggests that PMS is
Supporting Information).
also ineffective in Germany. Problems include the use of small sam-
The good news is that a revision of the GPSP is currently
ple sizes that make it impossible to detect rare adverse drug reac-
underway. A working group consisting of members of the MHLW,
tions (ADRs), the strict obligation of physicians to maintain
Pharmaceutical and Medical Devices Agency (PMDA) and industry
confidentiality towards sponsors, and the low rate (