Animals with tumors are recently treated by surgery, radiotherapy and chemotherapy, similarly to human medicine. To select optimal treatment for each tumor, many studies have been reported in veterinary medicine. Response rate and survival rate are generally used to evaluate each treatment. Prospective cohort study and experimental study, which have comparable internal control and few biases, are ideal, but take much time and cost. Realistically, most clinical studies are categorized to case report, case series study and retrospective study with external control. Our veterinary clinicians should create an environment to design ideal clinical study with statisticians.

Parmacoepidemiological studies that use medical information databases are powerful tools for post-marketing drug safety assessment. Recently, some medical information databases including claims data and electronic medical records data have become available in Japan. To conduct optimum drug safety assessment, researchers should comprehend the strengths and limitations of such databases. This paper deals with two important methodological topics in the field of pharmacoepidemiological ; causal inference and self-controlled methods. Also, pharmacoepidemiological studies may be important to find out rare serious adverse events of drugs, although few studies have focused on drug safety assessment in small animal medicine.

The domestic dog and cat have become integral to modern human family life. The Japanese dog and cat populations are estimated to be 10.35 and 9.96 million, with 15.1% and 10.1% of Japanese households being estimated to own at least one dog or cat respectively. With the integration of companion animals into human life, the number of owners who want to have their pets to receive proper veterinary medical care increased.

Knowing the pattern of disease occurrence of a defined animal population and the risk factors that may affect it, is important to maintain the health and prevent diseases in companion animals. It provides useful information that can be utilized by veterinary practitioners to guide diagnostic decision making, by breeders when planning breeding programs and by owner when acquiring a new pet. Despite substantial medical advances at the clinical level on diagnosis and treatment of diseases in individual animals, there is a shortage of epidemiological information in companion animals at population level.

　　In recent years, epidemiological analysis of dog diseases have been conducted using data from referral hospitals, and data from veterinary primary-care practices and veterinary practices. The data from referral hospitals are accurate in regard to diagnosis but they have no information about the total population at risk and the possible selection bias when only cases are referred to them. The data from primary-care practices and veterinary practices are more representative of the national dog population than those from referral hospitals, but they have a selection bias when a large proportion of dogs are not registered with practices or when the practices participating in the study are not representative of the overall veterinary practice structure. Pet insurance data have been used for research purposes since the 1970’s with increasing frequency during the last 15 years. Researchers have shown interest in pet insurance data because, although diagnostic information on insured animals may be inaccurate, they contain sound information relating to breed, sex and age of both diseased and healthy animals in the background population with less selection bias compared with data from referral hospitals and veterinary primary-care practices, and the insured population can be followed from enrolment to termination of coverage. Recently, epidemiological analysis by breed sex, age, and habitat has been conducted using data of insured dogs on mortality; mammary tumors, atopic dermatitis, lymphoma and bone tumors.

Although advances in veterinary medicine have led to a decrease in morbidity and have extended the life of pets, an evaluation of practice by a client not only depends on objective outcomes but it also involves subjective recognition. Thus, a veterinarian should also consider the psychosocial aspects of clients.

Grief derived from the loss of a pet is a well-known problem in this field. Therefore, there is a need for scales to quantify grief for the purpose of epidemiological research. Additionally, longitudinal assessments should be planned because short-term symptoms might be a part of the normal mourning. In that way, our study revealed that about half of the bereaved clients suffered from psychological symptoms two and four months after the death of their pet, indicating the need for psychiatric intervention for major depression.

Epidemiological studies on grief related to the loss of a pet would help develop preventive measures. However, we need to remember that the symptoms of a bereaved client may worsen owing to the stigma attached to psychiatric illnesses. is attached the stigma, it would aggravate the symptoms.

本研究は，タンザニア国アルーシャ市のビール・バーで販売されている「ニャマチョマ」と呼ばれる焼き牛肉の，カンピロバクター交差汚染によるリスクを評価することを目的として実施された。

2010年に，生牛肉と焼き牛肉における好熱性カンピロバクターの汚染率と最確数推定のため，層化無作為抽出した精肉店30件と焼き肉を提供するビール・バー30件を訪問した。さらに，カンピロバクターは一般的に家禽類でより多く保菌されているので，交差汚染の状態を観察するため焼き鶏肉を提供するバー10件を有意抽出して調査した。各精肉店とバーで一点の肉を収集し，質問票と直接観察により，販売ならびに衛生状況を評価した。リスクモデルを統計ソフトRで作成し，モンテカルロシミュレーションにより，アルーシャにおけるバーの客と成人男性集団における疾病発生率を推定した。

フィールド調査では，唯一焼き鶏肉サンプル一点からCampylobacter coli が検出され，その最確数は0.37/g（95% CI : 0.12—1.08）であった。アルーシャ市全体で，利用客における一日当たりカンピロバクター症発生数は0.15人（95% CI : 0.02—0.95）と推定された。利用客およびアルーシャ市成人男性における年間発生率は，千人当たりそれぞれ12.4人（95% CI : 1.2—83.6）と0.6人（95% CI : 0.06—4.0）と推定された。最も結果に影響を与える因子は焼き牛肉における好熱性カンピロバクター汚染率であり，次に最確数であった。多くのバー経営者（26/40, 65%）は異なる動物種由来の肉を提供しており，鶏肉と牛肉間での交差汚染は起こりやすい状況であった。ほぼ半数（18/39, 46%）の経営者は生肉と焼き肉に同じ包丁を使用していた。経営者の半数（20/40, 50%）が調理衛生研修を受講していたにも関わらず，研修の受講歴と交差汚染の主な原因と考えられる生と焼き肉の包丁の使い分けの実践とは統計学的関連性がなかった（x²＝0.22, df＝1, p＝0.6）。

結論として，焼き牛肉喫食によるカンピロバクター症は公衆衛生学的重要性が低いことが分かった。しかしながら，他の菌種のリスクを考えると，ビール・バーの食品による食中毒リスクをさらに低減するには，調理衛生研修内容の改善が望まれる。

FSCJ conducted assessments on human health risk associated with Bovine Spongiform Encephalopathy (BSE) in relation to the proposed revision of the domestic measures, in response to requests from the Ministry of Health, Labour and Welfare (MHLW). Using the findings obtained from published documents and reference materials /documents submitted by the MHWL, FSCJ first completed the risk assessment of the limit of cattle age for BSE testing, and subsequently will conduct the assessment of definition of Specified Risk Materials (SRMs).

For the age limit, FSCJ thoroughly assessed the possible risk of development of human prion diseases including vCJD linked with BSE prions through consumption of cattle meat and offal in association with the proposed cease of BSE testing of healthy slaughtered cattle for human consumption.

As the result, regarding the limit of cattle age for BSE testing, a difference between the following two measures would be extremely small in their influences on the risk to human health. One is to test continuously healthy slaughtered cattle over 48 months of age for human consumption, and the other is to cease BSE testing of healthy slaughtered cattle for human consumption. Therefore, FSCJ considers that the risk to human health arisen from the change of the measure on BSE testing from the former to the latter is negligible.

In recent years, the wild animal issues became serious, and many related laws and regulations have been revised. In particular, as the number of deer and wild boars population increased, it became uncontrollable, so in 2014 the birds and animal protection law was revised and private enterprises and others were able to enter the capture project. However, it seems that these revisions will be effective for the time being.