Table 3 Identified gaps in research for each article

MacEwen et al. [7]

x

x

x

x

x

x

Simon et al. [10]

x

x

x

Some other known prognostic factors (e.g. tumour size and stage) lost their impact on treatment outcome because of small number of patients

Low number of dogs with stage I

Pena et al. [30]

x

x

x

x

No information on how type of surgery was selected for each case

Betz et al. [25]

x

x

x

Type of surgery was chosen based on surgeon’s preference (no standardization)

Uneven distribution of benign and malignant tumours when compared to other studies

Histopathology on local recurrences and metastases was only performed on a limited number of cases

Stratmann et al. [26]

x

x

No control group

Incomplete staging (status of local lymph node unknown)

No subgrouping according to tumour size (1, 2 and 3 cm tumours)

Misdorp and Hart [31]

x

x

x

x

x

x

x

Old staging system

Misdorp and Hart [28, 29]

x

x

x

x

x

x

Old staging system

High percentage of dirty margins

Allen and Mahaffey [27]

x

x

x

x

Low numbers of patients in each treatment group

Old staging system

Invasion into skin, muscle or body wall was incorporated into an older classification system and it was therefore not possible to separate dogs with ulcerated tumours from dogs with tumours fixated to underlying tissue

Wey et al. [24]

x

Itoh et al. [33]

x

x

x

x

x

x

Low number of malignant tumours

Low number of dogs eligible for survival analysis (14 dogs)

Chang et al. [22]

x

x

x

x

x

x

Small groups of dogs

Stages I, II and III were grouped so stage I cases could not be extracted

Dias et al. [32]

x

x

x

x

x

x

Unknow number of dogs within each treatment group